Well-tuned Simple Nets Excel on Tabular Datasets

Arlind Kadra Marius Lindauer

Department of Computer Science Institute for Information Processing
University of Freiburg Leibniz University Hannover
[email protected] [email protected]
arXiv:2106.11189v2 [cs.LG] 5 Nov 2021

Frank Hutter Josif Grabocka

Department of Computer Science Department of Computer Science
University of Freiburg & University of Freiburg
Bosch Center for Artificial Intelligence [email protected]
[email protected]

Abstract

Tabular datasets are the last “unconquered castle” for deep learning, with traditional
ML methods like Gradient-Boosted Decision Trees still performing strongly even
against recent specialized neural architectures. In this paper, we hypothesize that
the key to boosting the performance of neural networks lies in rethinking the joint
and simultaneous application of a large set of modern regularization techniques.
As a result, we propose regularizing plain Multilayer Perceptron (MLP) networks
by searching for the optimal combination/cocktail of 13 regularization techniques
for each dataset using a joint optimization over the decision on which regularizers
to apply and their subsidiary hyperparameters.
We empirically assess the impact of these regularization cocktails for MLPs in a
large-scale empirical study comprising 40 tabular datasets and demonstrate that
(i) well-regularized plain MLPs significantly outperform recent state-of-the-art
specialized neural network architectures, and (ii) they even outperform strong
traditional ML methods, such as XGBoost.

1 Introduction

In contrast to the mainstream in deep learning (DL), in this paper, we focus on tabular data, a
domain that we feel is understudied in DL. Nevertheless, it is of great relevance for many practical
applications, such as climate science, medicine, manufacturing, finance, recommender systems, etc.
During the last decade, traditional machine learning methods, such as Gradient-Boosted Decision
Trees (GBDT) [5], dominated tabular data applications due to their superior performance, and the
success story DL has had for raw data (e.g., images, speech, and text) stopped short of tabular data.
Even in recent years, the existing literature still gives mixed messages on the state-of-the-art status of
deep learning for tabular data. While some recent neural network methods [1, 46] claim to outperform
GBDT, others confirm that GBDT are still the most accurate method on tabular data [48, 26]. The
extensive experiments on 40 datasets we report indeed confirm that recent neural networks [1, 46, 11]
do not outperform GBDT when the hyperparameters of all methods are thoroughly tuned.
We hypothesize that the key to improving the performance of neural networks on tabular data lies in
exploiting the recent DL advances on regularization techniques (reviewed in Section 3), such as data
augmentation, decoupled weight decay, residual blocks and model averaging (e.g., dropout or snapshot
ensembles), or on learning dynamics (e.g., look-ahead optimizer or stochastic weight averaging).

35th Conference on Neural Information Processing Systems (NeurIPS 2021), Sydney, Australia.
Indeed, we find that even plain Multilayer Perceptrons (MLPs) achieve state-of-the-art results when
regularized by multiple modern regularization techniques applied jointly and simultaneously.
Applying multiple regularizers jointly is already a common standard for practitioners, who routinely
mix regularization techniques (e.g. Dropout with early stopping and weight decay). However, the
deeper question of “Which subset of regularizers gives the largest generalization performance on
a particular dataset among dozens of available methods?” remains unanswered, as practitioners
currently combine regularizers via inefficient trial-and-error procedures. In this paper, we provide
a simple, yet principled answer to that question, by posing the selection of the optimal subset of
regularization techniques and their inherent hyperparameters, as a joint search for the best combination
of MLP regularizers for each dataset among a pool of 13 modern regularization techniques and their
subsidiary hyperparameters (Section 4).
From an empirical perspective, this paper is the first to provide compelling evidence that well-
regularized neural networks (even simple MLPs!) indeed surpass the current state-of-the-art models
in tabular datasets, including recent neural network architectures and GBDT (Section 6). In fact, the
performance improvements are quite pronounced and highly significant.1 We believe this finding to
potentially have far-reaching implications, and to open up a garden of delights of new applications on
tabular datasets for DL.
Our contributions are as follows:

1. We demonstrate that modern DL regularizers (developed for DL applications on raw data,

such as images, speech, or text) also substantially improve the performance of deep multi-
layer perceptrons on tabular data.
2. We propose a simple, yet principled, paradigm for selecting the optimal subset of regulariza-
tion techniques and their subsidiary hyperparameters (so-called regularization cocktails).
3. We demonstrate that these regularization cocktails enable even simple MLPs to outperform
both recent neural network architectures, as well as traditional strong ML methods, such as
GBDT, on tabular data. Specifically, we are the first to show neural networks to significantly
(and substantially) outperform XGBoost in a fair, large-scale experimental study.

2 Related Work on Deep Learning for Tabular Data

Recently, various neural architectures have been proposed for improving the performance of neural
networks on tabular data. TabNet [1] introduced a sequential attention mechanism for capturing
salient features. Neural oblivious decision ensembles (NODE [46]) blend the concept of hierarchical
decisions into neural networks. Self-normalizing neural networks [29] have neuron activations that
converge to zero mean and unit variance, which in turn, induces strong regularization and allows for
high-level representations. Regularization learning networks train a regularization strength on every
neural weight by posing the problem as a large-scale hyperparameter tuning scheme [48]. The recent
NET-DNF technique introduces a novel inductive bias in the neural structure corresponding to logical
Boolean formulas in disjunctive normal forms [26]. An approach that is often mistaken as deep
learning for tabular data is AutoGluon Tabular [11], which builds ensembles of basic neural networks
together with other traditional ML techniques, with its key contribution being a strong stacking
approach. We emphasize that some of these publications claim to outperform Gradient Boosted
Decision Trees (GDBT) [1, 46], while other papers explicitly stress that the neural networks tested
do not outperform GBDT on tabular datasets [48, 26]. In contrast, we do not propose a new kind of
neural architecture, but a novel paradigm for learning a combination of regularization methods.

3 An Overview of Regularization Methods for Deep Learning

Weight decay: The most classical approaches of regularization focused on minimizing the norms of
the parameter values, e.g., either the L1 [51], the L2 [52], or a combination of L1 and L2 known as
the Elastic Net [63]. A recent work fixes the malpractice of adding the decay penalty term before
1
In that sense, this paper adds to the growing body of literature on demonstrating the sensitivity of modern
ML methods to hyperparameter settings [19], demonstrating that proper hyperparameter tuning can yield
substantial improvements of modern ML methods [6], and demonstrating that even simple architectures can
obtain state-of-the-art performance with proper hyperparameter settings [38].

2
momentum-based adaptive learning rate steps (e.g., in common implementations of Adam [27]), by
decoupling the regularization from the loss and applying it after the learning rate computation [36].
Data Augmentation: Among the augmentation regularizers, Cut-Out [10] proposes to mask a subset
of input features (e.g., pixel patches for images) for ensuring that the predictions remain invariant to
distortions in the input space. Along similar lines, Mix-Up [60] generates new instances as a linear
span of pairs of training examples, while Cut-Mix [58] suggests super-positions of instance pairs
with mutually-exclusive pixel masks. A recent technique, called Aug-Mix [20], generates instances
by sampling chains of augmentation operations. On the other hand, the direction of reinforcement
learning (RL) for augmentation policies was elaborated by Auto-Augment [7], followed by a technique
that speeds up the training of the RL policy [34]. Recently, these complex and expensive methods
were superseded by simple and cheap methods that yield similar performance (RandAugment [8])
or even improve on it (TrivialAugment [41]). Last but not least, adversarial attack strategies (e.g.,
FGSM [17]) generate synthetic examples with minimal perturbations, which are employed in training
robust models [37].
Ensemble methods: Ensembled machine learning models have been shown to reduce variance and
act as regularizers [45]. A popular ensemble neural network with shared weights among its base
models is Dropout [49], which was extended to a variational version with a Gaussian posterior of
the model parameters [28]. As a follow-up, Mix-Out [32] extends Dropout by statistically fusing
the parameters of two base models. Furthermore, so-called snapshot ensembles [21] can be created
using models from intermediate convergence points of stochastic gradient descent with restarts [35].
In addition to these efficient ensembling approaches, ensembling independent classifiers trained in
separate training runs can yield strong performance (especially for uncertainty quantification), be it
based on independent training runs only differing in random seeds (deep ensembles [31]), training
runs differing in hyperparameter settings (hyperdeep ensembles, [55]), or training runs with different
neural architectures (neural ensemble search [59]).
Structural and Linearization: In terms of structural regularization, ResNet adds skip connections
across layers [18], while the Inception model computes latent representations by aggregating diverse
convolutional filter sizes [50]. A recent trend adds a dosage of linearization to deep models, where
skip connections transfer embeddings from previous less non-linear layers [18, 22]. Along similar
lines, the Shake-Shake regularization deploys skip connections in parallel convolutional blocks and
aggregates the parallel representations through affine combinations [15], while Shake-Drop extends
this mechanism to a larger number of CNN architectures [56].
Implicit: The last family of regularizers broadly encapsulates methods that do not directly propose
novel regularization techniques but have an implicit regularization effect as a virtue of their ‘modus
operandi’ [2]. The simplest such implicit regularization is Early Stopping [57], which limits overfitting
by tracking validation performance over time and stopping training when validation performance no
longer improves. Another implicit regularization method is Batch Normalization, which improves
generalization by reducing internal covariate shift [24]. The scaled exponential linear units (SELU)
represent an alternative to batch-normalization through self-normalizing activation functions [30]. On
the other hand, stabilizing the convergence of the training routine is another implicit regularization,
for instance by introducing learning rate scheduling schemes [35]. The recent strategy of stochastic
weight averaging relies on averaging parameter values from the local optima encountered along the
sequence of optimization steps [25], while another approach conducts updates in the direction of a
few ‘lookahead’ steps [61].

4 Regularization Cocktails for Multilayer Perceptrons

4.1 Problem Definition

A training set is composed of features X(Train) and targets y(Train) , while the test dataset is denoted
by X(Test) , y(Test) . A parametrized function f , i.e., a neural network, approximates the targets as
ŷ = f (X; θ), where the parameters

θ are trained to minimize a differentiable loss function
L
as arg minθ L y(Train) , f X(Train) ; θ . To generalize into minimizing L y(Test) , f (X(Test) ; θ , the
parameters of f are controlled with a regularization technique Ω that avoids overfitting to the
peculiarities of the training data. With a slight abuse of notation we denote f (X; Ω (θ; λ)) to be the
predictions of the model f whose parameters θ are optimized under the regime of the regularization

3
method Ω(·; λ), where λ ∈ Λ represents the hyperparameters of Ω. The training data is further
divided into two subsets as training and validation splits, the later denoted by X(Val) , y(Val) , such that
λ can be tuned on the validation loss via the following hyperparameter optimization objective:
λ∗ ∈ arg min L y(Val) , f X(Val) ; θλ ∗



, (1)
λ∈Λ
∗
∈ arg min L y(Train) , f (X(Train) ; Ω (θ; λ) .


s.t. θλ
θ

After finding the optimal (or in practice at least a well-performing) configuration λ∗ , we re-fit θ on
the entire training dataset, i.e., X(Train) ∪ X(Val) and y(Train) ∪ y(Val) .
While the search for optimal hyperparameters λ is an active field of research in the realm of
AutoML [23], still the choice of the regularizer Ω mostly remains an ad-hoc practice, where
practitioners select a few combinations among popular regularizers (Dropout, L2, Batch Nor-
malization, etc.). In contrast to prior studies, we hypothesize that the optimal regularizer is
a cocktail mixture of a large set of regularization methods, all being simultaneously applied
with different strengths (i.e., dataset-specific hyperparameters). Given a set of K regularizers
 (k)
K 


(Ω ·; λ(k) k=1 := Ω(1) ·; λ(1) , . . . , Ω(K) ·; λ(K) , each with its own hyperparameters
λ(k) ∈ Λ(k) , ∀k ∈ {1, . . . , K}, the problem of finding the optimal cocktail of regularizers is:

λ∗ ∈ ∗
L y(Val) , f X(Val) ; θλ



arg min (2)
λ:=(λ(1) ,...,λ(K) )∈(Λ(1) ,...,Λ(K) )
  n  oK 
∗
s.t.: θλ ∈ arg min L y(Train) , f X(Train) ; Ω(k) θ, λ(k)
θ k=1

The intuitive interpretation of Equation 2 is searching for the optimal hyperparameters λ (i.e.,
strengths) of the cocktail’s regularizers using the validation set, given that the optimal prediction
model parameters θ are trained under the regime of all the regularizers being applied jointly. We
stress that, for each regularizer, the hyperparameters λ(k) include a conditional hyperparameter
controlling whether the k-th regularizer is applied or skipped. The best cocktail might comprise only
a subset of regularizers.

4.2 Cocktail Search Space

To build our regularization cocktails we combine the 13 regularization methods listed in Table 1,
which represent the categories of regularizers covered in Section 3. The regularization cocktail’s
search space with the exact ranges for the selected regularizers’ hyperparameters is given in the same
table. In total, the optimal cocktail is searched in a space of 19 hyperparameters.
While we can in principle use any hyperparameter optimization method, we decided to use the
multi-fidelity Bayesian optimization method BOHB [12] since it achieves strong performance across
a wide range of computing budgets by combining Hyperband [33] and Bayesian Optimization [40],
and since BOHB can deal with the categorical hyperparameters we use for enabling or disabling
regularization techniques and the corresponding conditional structures. Appendix A describes the
implementation details for the deployed HPO method. Some of the regularization methods cannot be
combined, and we, therefore, introduce the following constraints to the proposed search space: (i)
Shake-Shake and Shake-Drop are not simultaneously active since the latter builds on the former; (ii)
Only one data augmentation technique out of Mix-Up, Cut-Mix, Cut-Out, and FGSM adversarial
learning can be active at once due to a technical limitation of the base library we use [62].

5 Experimental Protocol
5.1 Experimental Setup and Datasets

We use a large collection of 40 tabular datasets (listed in Table 9 of Appendix D). This includes 31
datasets from the recent open-source OpenML AutoML Benchmark [16]2 . In addition, we added
2
The remaining 8 datasets from that benchmark were too large to run effectively on our cluster.

4
Group Regularizer Hyperparameter Type Range Conditionality
BN BN-active Boolean {True, False} −
Implicit
SWA SWA-active Boolean {True, False} -
LA-active Boolean {True, False} −
LA Step size Continuous [0.5, 0.8] LA-active
Num. steps Integer [5, 10] LA-active
W. Decay WD-active Boolean {True, False} −
WD
Decay factor Continuous [10−5 , 0.1] WD-active
DO-active Boolean {True, False} −
Ensemble
{funnel, long funnel,
DO
Dropout shape Nominal diamond, hexagon, DO-active
brick, triangle, stairs}
Drop rate Continuous [0.0, 0.8] DO-active
SE SE-active Boolean {True, False} -
SC-active Boolean {True, False} −
SC
Structural MB choice Nominal {SS, SD, Standard} SC-active
SD Max. probability Continuous [0.0, 1.0] SC-active ∧ MB choice = SD
SS - - - SC-active ∧ MB choice = SS
− Augment Nominal {MU, CM, CO, AT, None} −
Augmentation MU Mix. magnitude Continuous [0.0, 1.0] Augment = MU
CM Probability Continuous [0.0, 1.0] Augment = CM
Probability Continuous [0.0, 1.0] Augment = CO
CO
Patch ratio Continuous [0.0, 1.0] Augment = CO
AT - - - Augment = AT

Table 1: The configuration space for the regularization cocktail regarding the explicit regularization
hyperparameters of the methods and the conditional constraints enabling or disabling them. (BN:
Batch Normalization, SWA: Stochastic Weight Averaging, LA: Lookahead Optimizer, WD: Weight
Decay, DO: Dropout, SE: Snapshot Ensembles, SC: Skip Connection, MB: Multi-branch choice,
SD: Shake-Drop, SS: Shake-Shake, MU: Mix-Up, CM: Cut-Mix, CO: Cut-Out, and AT: FGSM
Adversarial Learning)

9 popular datasets from UCI [3] and Kaggle that contain roughly 100K+ instances. Our resulting
benchmark of 40 datasets includes tabular datasets that represent diverse classification problems,
containing between 452 and 416 188 data points, and between 4 and 2 001 features, varying in terms
of the number of numerical and categorical features. The datasets are retrieved from the OpenML
repository [54] using the OpenML-Python connector [14] and split as 60% training, 20% validation,
and 20% testing sets. The data is standardized to have zero mean and unit variance where the statistics
for the standardization are calculated on the training split.
We ran all experiments on a CPU cluster, each node of which contains two Intel Xeon E5-2630v4
CPUs with 20 CPU cores each, running at 2.2GHz and a total memory of 128GB. We chose
the PyTorch library [43] as a deep learning framework and extended the AutoDL-framework
Auto-Pytorch [39, 62] with our implementations for the regularizers of Table 1. We provide the
code for our implementation at the following link: https://github.com/releaunifreiburg/
WellTunedSimpleNets.
To optimally utilize resources, we ran BOHB with 10 workers in parallel, where each worker had
access to 2 CPU cores and 12GB of memory, executing one configuration at a time. Taking into
account the dimensions D of the considered configuration spaces, we ran BOHB for at most 4 days,
or at most 40 × D hyperparameter configurations, whichever came first. During the training phase,
each configuration was run for 105 epochs, in accordance with the cosine learning rate annealing with
restarts (described in the following subsection). For the sake of studying the effect on more datasets,
we only evaluated a single train-val-test split. After the training phase is completed, we report the
results of the best hyperparameter configuration found, retrained on the joint train and validation set.

5.2 Fixed Architecture and Optimization Hyperparameters

In order to focus exclusively on investigating the effect of regularization we fix the neural architecture
to a simple multilayer perceptron (MLP) and also fix some hyperparameters of the general training

5
procedure. These fixed hyperparameter values, as specified in Table 4 of Appendix B.1, have been
tuned for maximizing the performance of an unregularized neural network on our dataset collection
(see Table 9 in Appendix D). We use a 9-layer feed-forward neural network with 512 units for each
layer, a choice motivated by previous work [42].
Moreover, we set a low learning rate of 10−3 after performing a grid search for finding the best
value across datasets. We use AdamW [36], which implements decoupled weight decay, and cosine
annealing with restarts [35] as a learning rate scheduler. Using a learning rate scheduler with restarts
helps in our case because we keep a fixed initial learning rate. For the restarts, we use an initial budget
of 15 epochs, with a budget multiplier of 2, following published practices [62]. Additionally, since
our benchmark includes imbalanced datasets, we use a weighted version of categorical cross-entropy
and balanced accuracy [4] as the evaluation metric.

5.3 Research Hypotheses and Associated Experiments

Hypothesis 1: Regularization cocktails outperform state-of-the-art deep learning architectures on

tabular datasets.
Experiment 1: We compare our well-regularized MLPs against the recently proposed deep learning
architectures Node [46] and TabNet [1]. Additionally, we compare against two versions
of AutoGluon Tabular [11], a version that features stacking and a version that addition-
ally includes hyperparameter optimization. Moreover, we add an unregularized version
of our MLP for reference, as well as a version of our MLP regularized with Dropout
(where the dropout hyperparameters are tuned on every dataset). Lastly, we also compare
against self-normalizing neural networks [29] by using the same MLP backbone as with our
regularization cocktails.
Hypothesis 2: Regularization cocktails outperform Gradient-Boosted Decision Trees (GBDTs), the
most commonly used traditional ML method and de-facto state-of-the-art for tabular data.
Experiment 2: We compare against three different implementations of GBDT: an implementation
from scikit-learn [44] and optimized by Auto-sklearn [13], the popular XGBoost [5], and
lastly, the recently proposed CatBoost [47].
Hypothesis 3: Regularization cocktails are time-efficient and achieve strong anytime results.
Experiment 3: We compare our regularization cocktails against XGBoost over time.

5.4 Experimental Setup for the Baselines

All baselines use the same train, validation, and test splits, the same seed, and the same HPO resources
and constraints as for our automatically-constructed regularization cocktails (4 days on 20 CPU cores
with 128GB of memory). After finding the best incumbent configuration, the baselines are refitted on
the union of the training and validation sets and evaluated on the test set. The baselines consist of
two recent neural architectures, two versions of AutoGluon Tabular with neural networks, and three
implementations of GBDT, as follows:

TabNet: This library does not provide an HPO algorithm by default; therefore, we also used BOHB
for this search space, with the hyperparameter value ranges recommended by the authors [1].
Node: This library does not offer an HPO algorithm by default. We performed a grid search among
the hyperparameter value ranges as proposed by the authors [46]; however, we faced multiple
memory and runtime issues in running the code. To overcome these issues we used the
default hyperparameters the authors used in their public implementation.
AutoGluon Tabular: This library constructs stacked ensembles with bagging among diverse neural
network architectures having various kinds of regularization [11]. The training of the
stacking ensemble of neural networks and its hyperparameter tuning are integrated into
the library. Hyperparameter optimization (HPO) is deactivated by default to give more
resources to stacking, but here we study AutoGluon based on either stacking or HPO (and
HPO actually performs somewhat better). While AutoGluon Tabular by default uses a broad
range of traditional ML techniques, here, in order to study it as a “pure” deep learning
method, we restrict it to only use neural networks as base learners.

6
ASK-GBDT: The GBDT implementation of scikit-learn offered by Auto-sklearn [13] uses SMAC
for HPO, and we used the default hyperparameter search space given by the library.
XGBoost: The original library [5] does not incorporate an HPO algorithm by default, so we used
BOHB for its HPO. We defined a search space for XGBoost’s hyperparameters following
the best practices by the community; we describe this in the Appendix B.2.
CatBoost: Like for XGBoost, the original library [47] does not incorporate an HPO algorithm, so
we used BOHB for its HPO, with the hyperparameter search space recommended by the
authors.

For in-depth details about the different baseline configurations with the exact hyperparameter search
spaces, please refer to Appendix B.2.

Dataset #Ins./#Feat. MLP MLP+D XGB. ASK-G. TabN. Node AutoGL. S MLP+C
anneal 898 / 39 84.131 86.916 85.416 90.000 84.248 20.000 80.000 89.270
kr-vs-kp 3196 / 37 99.701 99.850 99.850 99.850 93.250 97.264 99.687 99.850
arrhythmia 452 / 280 37.991 38.704 48.779 46.850 43.562 N/A 48.934 61.461
mfeat. 2000 / 217 97.750 98.000 98.000 97.500 97.250 97.250 98.000 98.000
credit-g 1000 / 21 69.405 68.095 68.929 71.191 61.190 73.095 69.643 74.643
vehicle 846 / 19 83.766 82.603 74.973 80.165 79.654 75.541 83.793 82.576
kc1 2109 / 22 70.274 72.980 66.846 63.353 52.517 55.803 67.270 74.381
adult 48842 / 15 76.893 78.520 79.824 79.830 77.155 78.168 80.557 82.443
walking. 149332 / 5 60.997 63.754 61.616 62.764 56.801 N/A 60.800 63.923
phoneme 5404 / 6 87.514 88.387 87.972 88.341 86.824 82.720 83.943 86.619
skin-seg. 245057 / 4 99.971 99.962 99.968 99.967 99.961 N/A 99.973 99.953
ldpa 164860 / 8 62.831 67.035 99.008 68.947 54.815 N/A 53.023 68.107
nomao 34465 / 119 95.917 96.232 96.872 97.217 95.425 96.217 96.420 96.826
cnae 1080 / 857 87.500 90.741 94.907 93.519 89.352 96.759 92.593 95.833
blood. 748 / 5 67.836 68.421 62.281 64.985 64.327 50.000 67.251 67.617
bank. 45211 / 17 78.076 83.145 72.658 72.283 70.639 74.607 79.483 85.993
connect. 67557 / 43 73.627 76.345 72.374 72.645 72.045 N/A 75.622 80.073
shuttle 58000 / 10 99.475 99.892 98.563 98.571 88.017 42.805 83.433 99.948
higgs 98050 / 29 67.752 66.873 72.944 72.926 72.036 N/A 73.798 73.546
australian 690 / 15 86.268 86.268 89.717 88.589 85.278 83.468 88.248 87.088
car 1728 / 7 97.442 99.690 92.376 100.000 98.701 46.119 99.675 99.587
segment 2310 / 20 94.805 94.589 93.723 93.074 91.775 90.043 91.991 93.723
fashion. 70000 / 785 90.464 90.507 91.243 90.457 89.793 N/A 91.336 91.950
jungle. 44819 / 7 97.061 97.237 87.325 83.070 73.425 N/A 93.017 97.471
numerai 96320 / 22 50.262 50.301 52.363 52.421 51.599 52.364 51.706 52.668
devnagari 92000 / 1025 96.125 97.000 93.310 77.897 94.179 N/A 97.734 98.370
helena 65196 / 28 16.836 23.983 21.994 21.144 19.032 N/A 27.115 27.701
jannis 83733 / 55 51.505 55.118 55.225 55.593 56.214 N/A 58.526 65.287
volkert 58310 / 181 65.081 66.996 64.170 63.428 59.409 N/A 70.195 71.667
miniboone 130064 / 51 90.639 94.099 94.024 94.137 62.173 N/A 94.978 94.015
apsfailure 76000 / 171 87.759 91.194 88.825 91.797 51.444 N/A 88.890 92.535
christine 5418 / 1637 70.941 70.756 74.815 74.447 69.649 73.247 74.170 74.262
dilbert 10000 / 2001 96.930 96.733 99.106 98.704 97.608 N/A 98.758 99.049
fabert 8237 / 801 63.707 64.814 70.098 70.120 62.277 66.097 68.142 69.183
jasmine 2984 / 145 78.048 76.211 80.546 78.878 76.690 80.053 80.046 79.217
sylvine 5124 / 21 93.070 93.363 95.509 95.119 83.595 93.852 93.753 94.045
dionis 416188 / 61 91.905 92.724 91.222 74.620 83.960 N/A 94.127 94.010
aloi 108000 / 129 92.331 93.852 95.338 13.534 93.589 N/A 97.423 97.175
ccfraud 284807 / 31 50.000 50.000 90.303 92.514 85.705 N/A 91.831 92.531
clickpred. 399482 / 12 63.125 64.367 58.361 58.201 50.163 N/A 54.410 64.280
Wins/Losses/Ties MLP+C vs . . . 35/5/0 30/8/2 26/12/2 29/11/0 38/2/0 19/2/0 30/9/1 -
Wilcoxon p-value MLP+C vs . . . 5.3 × 10−7 8.9 × 10−6 6 × 10−4 2.8 × 10−4 4.5 × 10−8 8.2 × 10−8 4 × 10−5 -

Table 2: Comparison of well-regularized MLPs vs. other methods in terms of balanced accuracy.
N/A values indicate a failure due to exceeding the cluster’s memory (24GB per process) or runtime
limits (4 days). The acronyms stand for MLP+D: MLP with Dropout, XGB.: XGBoost, ASK-G.:
GBDT by Auto-sklearn, AutoGL. S: Autogluon with stacking enabled, TabN.: TabNet and MLP+C:
our MLP regularized by cocktails.

6 Experimental Results

We present the comparative results of our MLPs regularized with the proposed regularization cocktails
(MLP+C) against ten baselines (descriptions in Section 5.4): (a) two state-of-the-art architectures
(NODE, TabN.); (b) two AutoGluon Tabular variants with neural networks that features stacking
(AutoGL. S) and additionally HPO (AutoGL. HPO); (c) three Gradient-Boosted Decision Tree

7
 0.8
0.8
0.6
MLP + C Error Rate

MLP + C Error Rate

MLP + C Error Rate
0.6
0.6
0.4 0.4 0.4
0.2 0.2 0.2

0.0 0.2 0.4 0.6 0.00 0.25 0.50 0.75 0.00 0.25 0.50 0.75
AutoGL. S Error Rate XGB. Error Rate ASK-G. Error Rate
Figure 1: Comparison of our proposed dataset-specific cocktail (MLP+C) against the top three
baselines. Each dot in the plot represents a dataset, the y-axis our method’s errors and the x-axis the
baselines’ errors.

(GBDT) implementations (XGB., ASK-G., and CatBoost); (d) as well as three reference MLPs
(unregularized (MLP), and regularized with Dropout (MLP+D) [49] or SELU (MLP+SELU) [30]).
Table 2 shows the comparison against a subset
of the baselines, while the full detailed results Ranks
involving all the remaining baselines are located 9 8 7 6 5 4 3 2 1
in Table 13 in the appendix. It is worth re-
emphasizing that the hyperparameters of all the TabN. + ES 7.0128 2.2125 MLP + C
TabN. 6.7000 3.4375AutoGL. HPO
presented baselines (except the unregularized NODE 6.1905 3.6750 AutoGL. S
MLP, which has no hyperparameters and Au- MLP 5.5375 4.1750 MLP + D

toGL. S) are carefully tuned on a validation set

MLP + SELU 4.7000

as detailed in Section 5 and the appendices ref- (a) CD of MLP+C vs. neural networks
erenced therein. The table entries represent the Ranks
test sets’ balanced accuracies achieved over the 5 4 3 2 1
described large-scale collection of 40 datasets.
Figure 1 visualizes the results showing substan- XGB. + ES 3.8000 2.0750 MLP + C
CatBoost 3.2051 2.7875 XGB.
tial improvements for our method. ASK-G. 3.0875
To assess the statistical significance, we ana- (b) CD of MLP+C vs. GBDT
lyze the ranks of the classification accuracies Ranks
across the 40 datasets. We use the Critical 8 7 6 5 4 3 2 1
Difference (CD) diagram of the ranks based
on the Wilcoxon significance test, a standard TabN. 6.7000 2.4625 MLP + C
metric for comparing classifiers across multi- NODE 6.1905 3.4500AutoGL. HPO
CatBoost 4.3590 3.9000 AutoGL. S
ple datasets [9]. The overall empirical compar- ASK-G. 4.0875 3.9250 XGB.
ison of the elaborated methods is given in Fig-
ure 2. The analysis of neural network baselines (c) CD of MLP+C vs. all baselines
in Subplot 2a reveals a clear statistical signifi-
cance of the regularization cocktails against the Figure 2: Critical difference diagrams with a
other methods. Apart from AutoGluon (both Wilcoxon significance analysis on 40 datasets.
versions), the other neural architectures are not Connected ranks via a bold bar indicate that perfor-
competitive even against an MLP regularized mances are not significantly different (p > 0.05).
only with Dropout and optimized with our stan-
dard, fixed training pipeline of Adam with co-
sine annealing. To be even fairer to the weaker
baselines (TabNet and Node) we tried boosting them by adding early stopping (indicated with "+ES"),
but their rank did not improve. Overall, the large-scale experimental analysis shows that Hypothesis 1
in Section 5.3 is validated: well-regularized simple deep MLPs outperform specialized neural
architectures.
Next, we analyze the empirical significance of our well-regularized MLPs against the GBDT im-
plementations in Figure 2b. The results show that our MLPs outperform all three GBDT variants
(XGBoost, auto-sklearn, and CatBoost) with a statistically significant margin. We added early stop-
ping ("+ES") to XGBoost, but it did not improve its performance. Among the GBDT implementations,
XGBoost without early stopping has a non-significant margin over the GBDT version of auto-sklearn

8
  *URXSHG&RFNWDLO,QJUHGLHQWV



)UHTXHQFLHV
SE BN WD CM DO LA CO
72.5% 62.5% 57.5% 40% 35% 32.5% 30% 


BN & SE SE & WD BN & WD
50% 40% 37.5%  O
OLFLW FD\ LQJ XUD WLRQ
,PS KW'H $YHUDJ 6WUXFW PHQWD
H LJ O  XJ
: 0RGH D$
'DW

Figure 3: Left: Cocktail ingredients occurring in at least 30% of the datasets. Right: Clustered
histogram (union of member occurrences) with the acronyms from Table 1. Implicit: {BN, LA,
SWA}, M. Averaging: {DO, SE}, Structural: {SC, SS, SD}, D. Augmentation: {MU, CM, CO, AT}.

as well as CatBoost. We conclude that well-regularized simple deep MLPs outperform GBDT,
which validates Hypothesis 2 in Section 5.3.
The final cumulative comparison in Figure 2c
provides a further result: none of the special-
Time (Hours) Wins Ties Losses p-value
ized previous deep learning methods (TabNet,
NODE, AutoGluon Tabular) outperforms GBDT 0.25 21 1 17 0.8561
significantly. To the best of our awareness, this 0.5 25 1 13 0.0145
paper is therefore the first to demonstrate that 1 24 1 14 0.0120
neural networks beat GBDT with a statistically 2 27 1 11 0.0006
significant margin over a large-scale experimen- 4 28 1 11 0.0006
tal protocol that conducts a thorough hyperpa- 8 28 1 11 0.0004
16 28 1 11 0.0005
rameter optimization for all methods.
32 29 1 10 0.0003
Figure 3 provides a further analysis on the most 64 30 1 9 0.0002
prominent regularizers of the MLP cocktails, 96 30 1 9 0.0002
based on the frequency with which our HPO
procedure selected the various regularization Table 3: Comparing the cocktails and XGBoost
methods for each dataset’s cocktail. In the left over different HPO budgets. The statistics are
plot, we show the frequent individual regulariz- based on the test performance of the incumbent
ers, while in the right plot the frequencies are configurations over all the benchmark datasets.
grouped by types of regularizers. The grouping reveals that a cocktail for each dataset often has at
least one ingredient from every regularization family (detailed in Section 3), highlighting the need for
jointly applying diverse regularization methods.
Lastly, Table 3 shows the efficiency of our regularization cocktails compared to XGBoost over increas-
ing HPO budgets. The descriptive statistics are calculated from the hyperparameter configurations
with the best validation performance for all datasets during the HPO search, however, taking their
respective test performances for comparison. A dataset is considered in the comparison only if the
HPO procedure has managed to evaluate at least one hyperparameter configuration for the cocktail
or baseline. As the table shows, our regularization cocktails achieve a better performance in only
15 minutes for the majority of datasets. After 30 minutes of HPO time, regularization cocktails are
statistically significantly better than XGBoost. As more time is invested, the performance gap with
XGBoost increases, and the results get even more significant; this is further visualized in the ranking
plot over time in Figure 4. Based on these results, we conclude that regularization cocktails are
time-efficient and achieve strong anytime results, which validates Hypothesis 3 in Section 5.3.

7 Conclusion
Summary. Focusing on the important domain of tabular datasets, this paper studied improvements
to deep learning (DL) by better regularization techniques. We presented regularization cocktails,
per-dataset-optimized combinations of many regularization techniques, and demonstrated that these
improve the performance of even simple neural networks enough to substantially and significantly

9
surpass XGBoost, the current state-of-the-art method for tabular datasets. We conducted a large-
scale experiment involving 13 regularization methods and 40 datasets and empirically showed that
(i) modern DL regularization methods developed in the context of raw data (e.g., vision, speech, text)
substantially improve the performance of deep neural networks on tabular data; (ii) regularization
cocktails significantly outperform recent neural networks architectures, and most importantly iii)
regularization cocktails outperform GBDT on tabular datasets.
Limitations. To comprehensively study basic
principles, we have chosen an empirical evalu-
ation that has many limitations. We only stud-
ied classification, not regression. We only used 1.7
somewhat balanced datasets (the ratio of the

Average Rank
minority class and the majority class is above 1.6
0.05). We did not study the regimes of extremely 1.5 MLP + C rank
few or extremely many data points (our smallest XGBoost rank
data set contained 452 data points, our largest 1.4
416 188 data points). We also did not study
datasets with extreme outliers, missing labels, 1.3
semi-supervised data, streaming data, and many
100 101 102
more modalities in which tabular data arises. Time (Hours)
An important point worth noticing is that the re-
cent neural network architectures (Section 5.4)
could also benefit from our regularization cock- Figure 4: Ranking plot comparing XGBoost and
tails, but integrating the regularizers into these regularization cocktails over time.
baseline libraries requires considerable coding
efforts.
Future Work. This work opens up the door for a wealth of exciting follow-up research. Firstly,
the per-dataset optimization of regularization cocktails may be substantially sped up by using meta-
learning across datasets [53]. Secondly, as we have used a fixed neural architecture, our method’s
performance may be further improved by using joint architecture and hyperparameter optimization.
Thirdly, regularization cocktails should also be tested under all the data modalities under “Limitations”
above. In addition, it would be interesting to validate the gain of integrating our well-regularized
MLPs into modern AutoML libraries, by combining them with enhanced feature preprocessing and
ensembling.
Take-away. Even simple neural networks can achieve competitive classification accuracies on
tabular datasets when they are well regularized, using dataset-specific regularization cocktails found
via standard hyperparameter optimization.

Societal Implications
Enabling neural networks to advance the state-of-the-art on tabular datasets may open up a garden
of delights in many crucial applications, such as climate science, medicine, manufacturing, and
recommender systems. In addition, the proposed networks can serve as a backbone for applications
of data science for social good, such as the realm of fair machine learning where the associated
data are naturally in a tabular form. However, there are also potential disadvantages in advancing
deep learning for tabular data. In particular, even though complex GBDT ensembles are also hard to
interpret, simpler traditional ML methods are much more interpretable than deep neural networks;
we therefore encourage research on interpretable deep learning on tabular data.

Acknowledgements
We acknowledge funding by the Robert Bosch GmbH, by the Eva Mayr-Stihl Foundation, the
MWK of the German state of Baden-Württemberg, the BrainLinks-BrainTools CoE, the European
Research Council (ERC) under the European Union’s Horizon 2020 programme, grant no. 716721,
and the German Federal Ministry of Education and Research (BMBF, grant RenormalizedFlows
01IS19077C). The authors acknowledge support by the state of Baden-Württemberg through bwHPC
and the German Research Foundation (DFG) through grant no INST 39/963-1 FUGG.

10
References
[1] S. Arik and T. Pfister. Tabnet: Attentive interpretable tabular learning. In AAAI Conference on
Artificial Intelligence, 2021.
[2] S. Arora, N. Cohen, W. Hu, and Y. Luo. Implicit regularization in deep matrix factorization. In
H. Wallach, H. Larochelle, A. Beygelzimer, F. d’Alche Buc, E. Fox, and R. Garnett, editors,
Advances in Neural Information Processing Systems 32, pages 7413–7424. Curran Associates,
Inc., 2019.
[3] A. Asuncion and D. Newman. Uci machine learning repository, 2007.
[4] K. Henning Brodersen, C., and E. Klaas Stephan J. M Buhmann. The balanced accuracy and its
posterior distribution. In 2010 20th International Conference on Pattern Recognition, pages
3121–3124. IEEE, 2010.
[5] T. Chen and C. Guestrin. XGBoost: A scalable tree boosting system. In Proceedings of the 22nd
ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages
785–794, 2016.
[6] Yutian Chen, Aja Huang, Ziyu Wang, Ioannis Antonoglou, Julian Schrittwieser, David Silver,
and Nando de Freitas. Bayesian optimization in alphago. CoRR, abs/1812.06855, 2018.
[7] E. Cubuk, B. Zoph, D. Mane, and V.Vasudevanand Q. Le. Autoaugment: Learning augmentation
strategies from data. In The IEEE Conference on Computer Vision and Pattern Recognition
(CVPR), June 2019.
[8] Ekin D. Cubuk, Barret Zoph, Jonathon Shlens, and Quoc V. Le. Randaugment: Practical
automated data augmentation with a reduced search space. In Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, June 2020.
[9] J. Demšar. Statistical comparisons of classifiers over multiple data sets. J. Mach. Learn. Res.,
7:1–30, December 2006.
[10] T. Devries and G. Taylor. Improved regularization of convolutional neural networks with cutout.
ArXiv, abs/1708.04552, 2017.
[11] N. Erickson, J. Mueller, A. Shirkov, H. Zhang, P. Larroy, M. Li, and A. Smola. Autogluon-
tabular: Robust and accurate automl for structured data. CoRR, abs/2003.06505, 2020.
[12] S. Falkner, A.Klein, and F. Hutter. BOHB: Robust and efficient hyperparameter optimization
at scalae. In Proceedings of the 35th International Conference on Machine Learning (ICML
2018), pages 1436–1445, July 2018.
[13] M. Feurer, A. Klein, K. Eggensperger, J. Springenberg, M. Blum, and F. Hutter. Efficient and
robust automated machine learning. In Proceedings of the 28th International Conference on
Neural Information Processing Systems - Volume 2, page 2755–2763. MIT Press, 2015.
[14] M. Feurer, JN. van Rijn, A. Kadra, P. Gijsbers, N. Mallik, S. Ravi, A. Müller, J. Vanschoren, and
F. Hutter. Openml-python: an extensible python api for openml. Journal of Machine Learning
Research, 22(100):1–5, 2021.
[15] X. Gastaldi. Shake-shake regularization of 3-branch residual networks. In 5th International
Conference on Learning Representations, ICLR. OpenReview.net, 2017.
[16] P. Gijsbers, E. LeDell, S. Poirier, J. Thomas, B. Bischl, and J. Vanschoren. An open source
automl benchmark. arXiv preprint arXiv:1907.00909 [cs.LG], 2019. Accepted at AutoML
Workshop at ICML 2019.
[17] I. Goodfellow, J. Shlens, and C. Szegedy. Explaining and harnessing adversarial examples. In
3rd International Conference on Learning Representations, ICLR, 2015.
[18] K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. In 2016
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778, 2016.

11
[19] P. Henderson, R. Islam, P. Bachman, J. Pineau, D. Precup, and D. Meger. Deep reinforcement
learning that matters. In S. McIlraith and K. Weinberger, editors, Proceedings of the Conference
on Artificial Intelligence (AAAI’18). AAAI Press, 2018.
[20] D. Hendrycks, N. Mu, E. Cubuk, B. Zoph, J. Gilmer, and B. Lakshminarayanan. Augmix:
A simple method to improve robustness and uncertainty under data shift. In International
Conference on Learning Representations, 2020.
[21] G. Huang, Y. Li, G. Pleiss, Z. Liu, J. Hopcroft, and K. Weinberger. Snapshot Ensembles: Train
1, Get M for Free. International Conference on Learning Representations, November 2017.
[22] G. Huang, Z. Liu, L. van der Maaten, and K. Weinberger. Densely connected convolutional
networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,
2017.
[23] F. Hutter, L. Kotthoff, and J. Vanschoren, editors. Automated Machine Learning: Methods,
Systems, Challenges. Springer, 2019. In press, available at http://automl.org/book.
[24] S. Ioffe and C. Szegedy. Batch normalization: Accelerating deep network training by reducing
internal covariate shift. In F. Bach and D. Blei, editors, Proceedings of the 32nd International
Conference on Machine Learning, volume 37 of Proceedings of Machine Learning Research,
pages 448–456. PMLR, 07–09 Jul 2015.
[25] P. Izmailov, D. Podoprikhin, T. Garipov, D. Vetrov, and A. Wilson. Averaging weights leads
to wider optima and better generalization. In Proceedings of the Thirty-Fourth Conference on
Uncertainty in Artificial Intelligence, UAI, pages 876–885. AUAI Press, 2018.
[26] L. Katzir, G. Elidan, and R. El-Yaniv. Net-{dnf}: Effective deep modeling of tabular data. In
International Conference on Learning Representations, 2021.
[27] D. Kingma and J. Ba. Adam: A method for stochastic optimization. In International Conference
on Learning Representations (ICLR), 2015.
[28] D. Kingma, T. Salimans, and M. Welling. Variational dropout and the local reparameterization
trick. In Proceedings of the 28th International Conference on Neural Information Processing
Systems - Volume 2, NIPS’15, page 2575–2583. MIT Press, 2015.
[29] G. Klambauer, T. Unterthiner, A. Mayr, and S. Hochreiter. Self-normalizing neural networks.
In Proceedings of the 31st international conference on neural information processing systems,
pages 972–981, 2017.
[30] Günter Klambauer, Thomas Unterthiner, Andreas Mayr, and Sepp Hochreiter. Self-normalizing
neural networks. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan,
and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 30. Curran
Associates, Inc., 2017.
[31] Balaji Lakshminarayanan, Alexander Pritzel, and Charles Blundell. Simple and scalable
predictive uncertainty estimation using deep ensembles. In Proceedings of the 31st International
Conference on Neural Information Processing Systems, NIPS’17, page 6405–6416, Red Hook,
NY, USA, 2017. Curran Associates Inc.
[32] C. Lee, K. Cho, and W. Kang. Mixout: Effective regularization to finetune large-scale pretrained
language models. In International Conference on Learning Representations, 2020.
[33] L. Li, K. Jamieson, G. DeSalvo, A. Rostamizadeh, and A. Talwalkar. Hyperband: A novel
bandit-based approach to hyperparameter optimization. J. Mach. Learn. Res., 18(1):6765–6816,
January 2017.
[34] S. Lim, I. Kim, T. Kim, C. Kim, and S. Kim. Fast autoaugment. In H. Wallach, H. Larochelle,
A. Beygelzimer, F. d Alche-Buc, E. Fox, and R. Garnett, editors, Advances in Neural Information
Processing Systems 32, pages 6665–6675. Curran Associates, Inc., 2019.
[35] I. Loshchilov and F. Hutter. Sgdr: Stochastic gradient descent with warm restarts. In In-
ternational Conference on Learning Representations (ICLR) 2017 Conference Track, April
2017.

12
[36] I. Loshchilov and F. Hutter. Decoupled weight decay regularization. In International Conference
on Learning Representations, 2019.
[37] A. Madry, A. Makelov, L. Schmidt, D. Tsipras, and A. Vladu. Towards deep learning models
resistant to adversarial attacks. In International Conference on Learning Representations, 2018.
[38] Gábor Melis, Chris Dyer, and Phil Blunsom. On the state of the art of evaluation in neural
language models. In International Conference on Learning Representations, 2018.
[39] H. Mendoza, A. Klein, M. Feurer, J. Tobias Springenberg, M. Urban, M. Burkart, M. Dippel,
M. Lindauer, and F. Hutter. Towards automatically-tuned deep neural networks. In F. Hutter,
L. Kotthoff, and J. Vanschoren, editors, AutoML: Methods, Sytems, Challenges, chapter 7, pages
141–156. Springer, December 2019.
[40] J. Mockus. Application of bayesian approach to numerical methods of global and stochastic
optimization. J. Glob. Optim., 4(4):347–365, 1994.
[41] Samuel G. Müller and Frank Hutter. Trivialaugment: Tuning-free yet state-of-the-art data
augmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision
(ICCV), pages 774–782, October 2021.
[42] A. Emin Orhan and X. Pitkow. Skip connections eliminate singularities. arXiv preprint
arXiv:1701.09175, 2017.
[43] A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen Z. Lin,
N. Gimelshein, L. Antiga, et al. Pytorch: An imperative style, high-performance deep learning
library. In Advances in neural information processing systems, pages 8026–8037, 2019.
[44] F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel,
P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher,
M. Perrot, and E. Duchesnay. Scikit-learn: Machine learning in Python. Journal of Machine
Learning Research, 12:2825–2830, 2011.
[45] R. Polikar. Ensemble Learning. In C. Zhang and Y. Ma, editors, Ensemble Machine Learning:
Methods and Applications, pages 1–34. Springer US, 2012.
[46] S. Popov, S. Morozov, and A. Babenko. Neural oblivious decision ensembles for deep learning
on tabular data. In International Conference on Learning Representations, 2020.
[47] L. Prokhorenkova, G. Gusev, A. Vorobev, AV. Dorogush, and A. Gulin. Catboost: unbiased
boosting with categorical features. Advances in Neural Information Processing Systems, 31,
2018.
[48] I. Shavitt and E. Segal. Regularization learning networks: Deep learning for tabular datasets. In
Proceedings of the 32nd International Conference on Neural Information Processing Systems,
page 1386–1396. Curran Associates Inc., 2018.
[49] N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov. Dropout: A
simple way to prevent neural networks from overfitting. J. Mach. Learn. Res., 15(1):1929–1958,
January 2014.
[50] C. Szegedy, S. Ioffe, V. Vanhoucke, and A. Alemi. Inception-v4, inception-resnet and the impact
of residual connections on learning. In Thirty-first AAAI conference on artificial intelligence,
2017.
[51] R. Tibshirani. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical
Society (Series B), 58:267–288, 1996.
[52] A. Tikhonov. On the stability of inverse problems. In Doklady Akademii Nauk SSSR, 1943.
[53] J. Vanschoren. Meta-learning. In F. Hutter, L. Kotthoff, and J. Vanschoren, editors, Automated
Machine Learning - Methods, Systems, Challenges, The Springer Series on Challenges in
Machine Learning, pages 35–61. Springer, 2019.

13
[54] J. Vanschoren, J. Van Rijn, B. Bischl, and L. Torgo. Openml: networked science in machine
learning. ACM SIGKDD Explorations Newsletter, 15(2):49–60, 2014.
[55] Florian Wenzel, Jasper Snoek, Dustin Tran, and Rodolphe Jenatton. Hyperparameter ensembles
for robustness and uncertainty quantification. In H. Larochelle, M. Ranzato, R. Hadsell, M. F.
Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems, volume 33,
pages 6514–6527. Curran Associates, Inc., 2020.
[56] Y. Yamada, M. Iwamura, and K. Kise. Shakedrop regularization, 2018.
[57] Y. Yao, L. Rosasco, and A. Caponnetto. On early stopping in gradient descent learning.
Constructive Approximation, 26(2):289–315, August 2007.
[58] S. Yun, D. Han, S. Joon, S. Chun, J. Choe, and Y. Yoo. Cutmix: Regularization strategy to train
strong classifiers with localizable features. In International Conference on Computer Vision
(ICCV), 2019.
[59] Sheheryar Zaidi, Arber Zela, Thomas Elsken, Chris Holmes, Frank Hutter, and Yee Whye Teh.
Neural ensemble search for uncertainty estimation and dataset shift. In Advances in Neural
Information Processing Systems, 2021.
[60] H. Zhang, M. Cisse, Y. Dauphin, and D. Paz. mixup: Beyond empirical risk minimization. In
International Conference on Learning Representations, 2018.
[61] M. Zhang, J. Lucas, J. Ba, and G. Hinton. Lookahead optimizer: k steps forward, 1 step back.
In H. Wallach, H. Larochelle, A. Beygelzimer, F. Buc, E. Fox, and R. Garnett, editors, Advances
in Neural Information Processing Systems 32: Annual Conference on Neural Information
Processing Systems 2019, pages 9593–9604, 2019.
[62] L. Zimmer, M. Lindauer, and F. Hutter. Auto-pytorch tabular: Multi-fidelity metalearning for
efficient and robust autodl. IEEE TPAMI, 2021. IEEE Early Access.
[63] H. Zou and T. Hastie. Regularization and variable selection via the elastic net. Journal of the
royal statistical society: series B (statistical methodology), 67(2):301–320, 2005.

14
A Description of BOHB
BOHB [12] is a hyperparameter optimization algorithm that extends Hyperband [33] by sampling
from a model instead of sampling randomly from the hyperparameter search space.
Initially, BOHB performs random search and favors exploration. As it iterates and gets more
observations, it builds models over different fidelities and trades off exploration with exploitation to
avoid converging in bad regions of the search space. BOHB samples from the model of the highest
fidelity with a probability p and with 1 − p from random. A model is built for a fidelity only when
enough observations exist for that fidelity; by default, this limit is set to equal S + 1 observations,
where S is the dimensionality of the search space.
We used the original implementation of BOHB in HPBandSter3 [12]. For simplicity, we only used a
single fidelity level for BOHB, effectively using it as a blackbox optimization algorithm; we believe
that future work should revisit this choice.

B Configuration Spaces
B.1 Method implicit search space

Category Hyperparameter Type Range

Iterations multiplier Continuous {2.0}
Cosine Annealing
Max. iterations Integer {15}
Activation Nominal {ReLU}
Bias initialization Nominal {Yes}
Blocks in a group Integer {2}
Embeddings Nominal {One-Hot encoding}
Network
Number of groups Integer {2}
Resnet shape Nominal {Brick}
Type Nominal {Shaped-Resnet}
Units in a layer Integer {512}
Preprocessing Preprocessor Nominal {None}
Batch size Integer {128}
Imputation Nominal {Median}
Initialization method Nominal {Default}
Training Learning rate Continuous {10−3 }
Loss module Nominal {Weighted Cross-Entropy}
Normalization strategy Nominal {Standardize}
Optimizer Nominal {AdamW}
Scheduler Nominal {COS}
Seed Integer {11}
Table 4: The configuration space of the training and model architecture hyperparameters. All these
hyperparameters only have one value in their range, meaning they are fixed.

Table 4 presents the network architecture and the training pipeline choices used in all our experiments
for the individual regularizers and for the regularization cocktails.

B.2 Benchmark search space

For the experiments conducted in our work, we set up the search space and the individual configura-
tions of the state-of-the-art competitors used for the comparison as follows:

Auto-Sklearn. The estimator is restricted to only include GBDT, for the sake of fully comparing
against the algorithm as a baseline. We do not activate any preprocessing since our regularization
3
https://github.com/automl/HpBandSter

15
cocktails also do not make use of preprocessing algorithms in the pipeline. The time left is always
selected based on the time it took BOHB to find the hyperparameter with the best validation accuracy
from the start of the hyperparameter optimization phase. The ensemble size is kept to 1 since our
method only uses models from one training run, not multiple ones. The seed is set to 11 as it was
set in the experiments with the regularization cocktail, to obtain the same data splits. To keep the
comparison fair, there is no warm start for the initial configurations with meta-learning, since our
method also does not make use of meta-learning. Lastly, the number of parallel workers is set to 10,
to match the parallel resources that were given to the experiment with the regularization cocktails.
The search space of the hyperparameters is left to the default search space offered by Auto-Sklearn
which is shown in Table 5. We use version 0.10.0 of the library.

Hyperparameter Type Range

early_stopping Nominal {Off, Train, Valid}
l2 _regularization Continuous [1e − 10, 1]
learning_rate Continuous [0.01, 1]
max_leaf _nodes Integer [3, 2047]
min_samples_leaf Integer [1, 200]
nr_iterations_no_change Integer [1, 20]
validation_f raction Continuous [0.01, 0.4]

Table 5: The search space of the training and model hyperparameters for the gradient boosting
estimator of the Auto-Sklearn tool.

XGBoost. To have a well-performing configuration space for XGBoost we augmented the default
configuration spaces previously used in Auto-Sklearn4 with further recommended hyperparameters
and ranges from Amazon5 . In Table 6, we present a refined version of the configuration space
that achieves a better performance on the benchmark. We would like to note that we did not apply
One-Hot encoding to the categorical features for the experiment, since we observed better overall
results when the categorical features were ordinal encoded. Nevertheless, in Table 13 we ablate the
choice of encoding (+ENC) as a categorical hyperparameter (one-hot vs. ordinal) and also early
stopping (+ES). When early stopping is activated, the num_rounds is increased to 4000. All the
results can be found in Appendix D.

TabNet. For the search space of the TabNet model, we used the default hyperparameter ranges
suggested by the authors which were found to perform best in their experiments.
For our experiments with the TabNet and XGBoost models, we also used BOHB for hyperparameter
tuning, using the same parallel resources and limiting conditions as for our regularization cocktail.
In the above search spaces for the experiments with the XGBoost and TabNet models, we did not
include early stopping; however, we did actually run experiments with early stopping for both models,
and results did not improve. Lastly, for both experiments, we imputed missing values with the most
frequent strategy (the implementation we used did not accept the median strategy for categorical
value imputation).

AutoGluon. The library is configured to construct stacked ensembles with bagging among diverse
neural network architectures having various kinds of regularization with the preset = ’Best Quality’
to achieve the best predictive accuracy. Furthermore, we used the same seed as for our MLPs with
regularization cocktails to obtain the same dataset splits. We allowed AutoGluon to make use of early
stopping and additionally, we allowed feature preprocessing since different feature preprocessing
techniques are embedded in different model types, to allow for better overall performance. For all
4
https://github.com/automl/auto-sklearn/blob/v.0.4.2/autosklearn/pipeline/
components/classification/xgradient_boosting.py
5
https://docs.aws.amazon.com/sagemaker/latest/dg/xgboost-tuning.html We reduced the
size of some ranges since the ranges given at this website were too broad and resulted in poor performance.

16
 Hyperparameter Type Range Log scale
eta Continuous [0.001, 1] X
lambda Continuous [1e − 10, 1] X
alpha Continuous [1e − 10, 1] X
num_round Integer [1, 1000] -
gamma Continuous [0.1, 1] X
colsample_bylevel Continuous [0.1, 1] -
colsample_bynode Continuous [0.1, 1] -
colsample_bytree Continuous [0.5, 1] -
max_depth Integer [1, 20] -
max_delta_step Integer [0, 10] -
min_child_weight Continuous [0.1, 20] X
subsample Continuous [0.01, 1] -
Table 6: The hyperparameter search space for the XGBoost library.

Hyperparameter Type Values

na Integer {8, 16, 24, 32, 64, 128}
learning_rate Continuous {0.005, 0.01, 0.02, 0.025}
gamma Continuous {1.0, 1.2, 1.5, 2.0}
nsteps Integer {3, 4, 5, 6, 7, 8, 9, 10}
λsparse Continuous {0, 0.000001, 0.0001, 0.001, 0.01, 0.1}
batch_size Integer {256, 512, 1024, 2048, 4096, 8192, 16384, 32768}
virtual_batch_size Integer {256, 512, 1024, 2048, 4096}
decay_rate Continuous {0.4, 0.8, 0.9, 0.95}
decay_iterations Integer {500, 2000, 8000, 10000, 20000}
momentum Continuous {0.6, 0.7, 0.8, 0.9, 0.95, 0.98}

Table 7: The hyperparameter search space for TabNet.

the other training and hyperparameter settings, we used the library’s default6 following the explicit
recommendation of the authors on the efficacy of their proposed stacking without needing any
HPO [11].
We also investigate using HPO with AutoGluon and compare the results against the version that uses
stacking; the full detailed results can be found in Appendix D.

NODE. For our experiments with NODE we used the official implementation7 . In our initial
experiment iterations, we used the search space that was proposed by the authors [46]. However,
evaluating the search space proposed is infeasible, since the memory and run-time requirements of the
experiments are very high and cannot be satisfied within our cluster constraints. The high run-time
and memory issues are also noted by the authors in the official implementation.
To alleviate these problems, we used the default configuration suggested by the authors in the
examples, where num_layers = 2, total_tree_count = 1024 and tree_depth = 6. Lastly, we use
the same seed as for our experiment with the regularization cocktails to obtain the same data splits.

6
We used version 0.2.0 of the AutoGluon library
7
https://github.com/Qwicen/node

17
 Hyperparameter Type Range Log scale
−7
learning_rate Continuous [e , 1] X
random_strength Integer [1, 20] -
one_hot_max_size Integer [0, 25] -
num_round Integer [1, 4000] -
l2_leaf _reg Continuous [1, 10] X
bagging_temperature Continuous [0, 1] -
gradient_iterations Integer [1, 10] -

Table 8: The hyperparameter search space for the CatBoost library.

CatBoost. We use version 0.26 of the official library and we use the hyperparameter search space
that is recommended by the authors [47], provided in Table 8.

C Plots

C.1 Regularization Cocktail Performance

&RFNWDLO :' '2 %1 6' 66 6& 6:$ 6( &2 &0 08 /$ $7 31
             
&RFNWDLO

:LOFR[RQVLJQHGUDQNWHVWSYDOXH
S S S S S S S S S S S S S S
            
:' S S S S S S S S S S S S S
           
'2 S S S S S S S S S S S S 
          
%1 S S S S S S S S S S S
         
6' S S S S S S S S S S
        
66 S S S S S S S S S 
       
6& S S S S S S S S
      
6:$ S S S S S S S
     
6( S S S S S S 
    
&2 S S S S S
   
&0 S S S S
  
08 S S S
  
/$ S S

$7 S
31

Figure 5: Pairwise statistical significance and comparison. For every entry, the first row show-
cases the wins, draws and losses of the horizontal method with the vertical method on all datasets,
calculated on the test set; the second row presents the p-value for the statistical significance test.

To investigate the performance of our formulation, we compare plain MLPs regularized with only one
individual regularization technique at a time against the dataset-specific regularization cocktails. The
hyperparameters for all methods are tuned on the validation set and the best configuration is refitted
on the full training set. In Figure 5, we present the results of each pairwise comparison. The results
presented are calculated on the test set after the refit phase is completed on the best hyperparameter
configuration. The p-value is generated by performing a Wilcoxon signed-rank test. As can be
seen from the results, the regularization cocktail is the only method that has statistically significant
improvements compared to all other methods (with a p-value ≤ 0.001 in all cases). The detailed
results for all methods on every dataset are shown in Table 11.

18
C.2 Dataset-dependent optimal cocktails

To verify the necessity for dataset-specific regularization cocktails, we initially investigate the
best-found hyperparameter configurations to observe the occurrences of individual regularization
techniques. In Figure 6, we present the occurrences of every regularization method over all datasets.
The occurrences are calculated by analyzing the best-found hyperparameter configuration for each
dataset and observing the number of times the regularization method was chosen to be activated by
BOHB. As can be seen from Figure 6, there is no regularization method or combination that is always
chosen for every dataset.
Additionally, we compare our regularization cocktails against the top-5 frequently chosen regular-
ization techniques and the top-5 best performing regularization techniques. For the top-5 baselines,
the regularization techniques are activated and their hyperparameters are tuned on the validation set.
The results of the comparison as shown in Table 10 show that the cocktail outperforms both top-5
variants, indicating the need for dataset-specific regularization cocktails.

 2FFXUHQFHVRIWKH&RFNWDLO,QJUHGLHQWV


2FFXUHQFHV






$7
6:$
%1
'2
6&
66
6'
/$

6(
:'
&2
&0
08

Figure 6: Frequency of the regularization techniques. The occurrences of the individual regu-
larization techniques in the best hyperparameter configurations found by the cocktail across 40
datasets.

C.3 Learning rate as a hyperparameter

In the majority of our experiments, we keep a fixed initial learning rate to investigate in detail the
effect of the individual regularization techniques and the regularization cocktails. The learning rate
is set to a fixed value that achieves the best results across the chosen benchmark of datasets. To
investigate the role and importance of the learning rate in the regularization cocktail performance, we
perform an additional experiment, where, the learning rate is an additional hyperparameter that is
optimized individually for every dataset. The results as shown in Table 12, indicate that regularization
cocktails with a dynamic learning rate outperform the regularization cocktails with a fixed learning
rate in 21 out of 40 datasets, tie in 1 and lose in 18. However, the results are not statistically significant
with a p-value of 0.7 and do not indicate a clear region where the dynamic learning rate helps.

D Tables
In Table 9, we provide information about the datasets that are considered in our experiments. Con-
cretely, we provide descriptive statistics and the identifiers for every dataset. The identifier (the task
id) can be used to download the datasets from OpenML (http://www.openml.org) by using the
OpenML-Python connector [14].
Table 10 shows the results for the comparison between the Regularization Cocktail and the Top-5
cocktail variants. The results are calculated on the test set for all datasets, after retraining on the best
dataset-specific hyperparameter configuration.
Table 11 provides the results of all our experiments for the plain MLP baseline, the individual
regularization methods, and the regularization cocktail. All the results are calculated on the test set,

19
 Task Id Dataset Name Number of Instances Number of Features Majority Class Percentage Minority Class Percentage
233090 anneal 898 39 76.17 0.89
233091 kr-vs-kp 3196 37 52.22 47.78
233092 arrhythmia 452 280 54.20 0.44
233093 mfeat-factors 2000 217 10.00 10.00
233088 credit-g 1000 21 70.00 30.00
233094 vehicle 846 19 25.77 23.52
233096 kc1 2109 22 84.54 15.46
233099 adult 48842 15 76.07 23.93
233102 walking-activity 149332 5 14.73 0.61
233103 phoneme 5404 6 70.65 29.35
233104 skin-segmentation 245057 4 79.25 20.75
233106 ldpa 164860 8 33.05 0.84
233107 nomao 34465 119 71.44 28.56
233108 cnae-9 1080 857 11.11 11.11
233109 blood-transfusion 748 5 76.20 23.80
233110 bank-marketing 45211 17 88.30 11.70
233112 connect-4 67557 43 65.83 9.55
233113 shuttle 58000 10 78.60 0.02
233114 higgs 98050 29 52.86 47.14
233115 Australian 690 15 55.51 44.49
233116 car 1728 7 70.02 3.76
233117 segment 2310 20 14.29 14.29
233118 Fashion-MNIST 70000 785 10.00 10.00
233119 Jungle-Chess-2pcs 44819 7 51.46 9.67
233120 numerai28.6 96320 22 50.52 49.48
233121 Devnagari-Script 92000 1025 2.17 2.17
233122 helena 65196 28 6.14 0.17
233123 jannis 83733 55 46.01 2.01
233124 volkert 58310 181 21.96 2.33
233126 MiniBooNE 130064 51 71.94 28.06
233130 APSFailure 76000 171 98.19 1.81
233131 christine 5418 1637 50.00 50.00
233132 dilbert 10000 2001 20.49 19.13
233133 fabert 8237 801 23.39 6.09
233134 jasmine 2984 145 50.00 50.00
233135 sylvine 5124 21 50.00 50.00
233137 dionis 416188 61 0.59 0.21
233142 aloi 108000 129 0.10 0.10
233143 C.C.FraudD. 284807 31 99.83 0.17
233146 Click prediction 399482 12 83.21 16.79

Table 9: Datasets. The collection of datasets used in our experiments, combined with detailed
information for each dataset.

Task Id Cockt. Top-5 F Top-5 R Task Id Cockt. Top-5 F Top-5 R Task Id Cockt. Top-5 F Top-5 R
233090 89.27 89.71 88.54 233091 99.85 99.85 98.20 233092 61.46 59.94 57.21
233093 98.00 98.75 98.75 233088 74.64 71.43 74.76 233094 82.58 82.01 80.33
233096 74.38 78.03 73.96 233099 82.44 82.35 82.24 233102 63.92 62.21 54.10
233103 86.62 85.90 82.33 233104 99.95 99.96 99.85 233106 68.11 68.81 55.45
233107 96.83 96.67 96.59 233108 95.83 95.83 95.83 233109 67.62 67.32 68.20
233110 85.99 86.35 86.06 233112 80.07 79.57 77.49 233113 99.95 97.95 85.34
233114 73.55 73.25 72.06 233115 87.09 88.11 87.60 233116 99.59 100.00 98.20
233117 93.72 93.94 90.69 233118 91.95 91.83 91.59 233119 97.47 92.66 85.53
233120 52.67 52.49 51.70 233121 98.37 98.41 96.93 233122 27.70 28.82 28.09
233123 65.29 65.13 62.11 233124 71.67 70.87 66.06 233126 94.02 88.13 93.16
233130 92.53 96.24 95.89 233131 74.26 71.86 74.63 233132 99.05 98.95 98.55
233133 69.18 68.75 69.03 233134 79.22 78.21 77.71 233135 94.05 94.43 93.95
233137 94.01 94.33 92.43 233142 97.17 97.06 96.06
233146 64.28 64.53 63.28 233143 92.53 92.13 92.59

Table 10: Top-5 baselines. The test set performance for the Regularization Cocktail against the
Top-5 Most Frequent (Top-5 F) and the Top-5 Highest Ranks (Top-5 R) baselines.

after retraining on the best-found hyperparameter configurations. The evaluation metric used for the
performance is balanced accuracy.
Additionally, in Table 12, we provide the results of the regularization cocktails with a fixed learning
rate and with the learning rate being a hyperparameter optimized for every dataset.
Lastly, in Table 13, we present the remaining baselines from our experiments/ablations and their
final performances on every dataset. The results show the test set performances after the incumbent
configuration is refit.

20
 Task Id PN BN LA SE SWA SC AT SS SD MU CO CM WD DO Cocktail
233090 84.13 86.78 83.99 86.48 87.96 87.21 86.92 84.28 87.21 89.27 85.60 86.77 87.06 86.92 89.27
233091 99.70 99.85 99.70 99.70 99.55 100.00 99.85 99.85 99.69 99.85 99.55 99.85 99.85 99.85 99.85
233092 37.99 41.91 36.14 37.31 25.94 53.42 38.79 55.61 53.26 42.19 32.48 42.22 35.76 38.70 61.46
233093 97.75 98.50 96.00 97.75 69.25 98.25 97.25 97.25 98.25 98.00 98.00 97.75 98.00 98.00 98.00
233088 69.40 68.69 70.83 69.76 69.40 66.43 69.29 66.43 67.14 70.00 70.36 64.29 69.29 68.10 74.64
233094 83.77 83.17 84.36 84.39 83.36 80.82 83.17 83.20 81.98 83.77 81.47 78.65 83.20 82.60 82.58
233096 70.27 66.56 71.95 76.43 75.44 77.40 71.95 65.31 78.31 72.43 76.84 74.94 67.33 72.98 74.38
233099 76.89 77.92 75.95 78.23 76.38 78.38 76.75 75.56 78.61 78.67 82.56 82.23 76.99 78.52 82.44
233102 61.00 62.89 61.32 63.57 56.67 60.79 59.99 43.04 60.77 61.95 63.30 63.49 64.03 63.75 63.92
233103 87.51 87.02 88.25 87.03 87.22 85.90 87.99 87.64 85.90 87.12 87.26 86.59 86.74 88.39 86.62
233104 99.97 99.96 99.96 99.94 2.57 99.97 99.95 92.77 99.97 99.95 99.96 99.97 99.96 99.96 99.95
233106 62.83 68.90 62.46 65.70 62.16 61.85 61.89 44.63 62.05 66.29 65.43 64.99 66.50 67.04 68.11
233107 95.92 95.93 96.01 96.36 95.23 95.76 95.77 95.37 96.22 96.52 96.10 96.55 95.98 96.23 96.83
233108 87.50 91.20 85.65 87.96 50.00 93.98 92.59 94.91 94.44 94.44 93.06 95.37 91.67 90.74 95.83
233109 67.84 73.68 66.52 68.20 66.45 65.20 66.89 66.74 67.03 68.64 67.32 70.18 66.23 68.42 67.62
233110 78.08 72.58 72.70 83.40 66.93 72.74 74.12 70.16 74.76 74.09 85.71 85.76 72.34 83.14 85.99
233112 73.63 74.68 73.37 74.33 77.36 73.86 72.91 72.06 74.35 72.08 76.23 75.74 72.48 76.35 80.07
233113 99.47 99.89 99.92 99.87 55.86 98.11 99.46 90.60 98.11 99.94 99.92 99.91 99.88 99.89 99.95
233114 67.75 68.90 68.81 69.11 67.36 68.08 67.44 67.70 68.56 68.59 71.93 73.13 67.80 66.87 73.55
233115 86.27 85.79 88.73 86.44 87.26 87.74 88.39 87.74 88.39 88.73 88.25 88.90 87.91 86.27 87.09
233116 97.44 100.00 96.79 97.44 87.35 99.47 99.14 97.46 99.69 99.37 97.64 99.04 97.44 99.69 99.59
233117 94.81 92.86 93.51 93.51 90.48 93.72 92.86 92.64 93.72 93.51 93.07 93.72 93.94 94.59 93.72
233118 90.46 90.86 90.73 90.75 81.72 89.91 90.69 86.69 90.06 91.11 91.09 91.88 90.70 90.51 91.95
233119 97.06 93.76 97.79 96.08 92.15 87.83 97.16 87.08 87.68 98.14 96.50 97.51 97.33 97.24 97.47
233120 50.26 50.95 51.29 50.50 51.63 50.92 50.17 50.23 51.00 50.72 52.35 52.10 50.41 50.30 52.67
233121 96.12 97.83 96.45 96.74 92.40 95.31 96.34 91.38 95.15 97.52 97.88 97.80 96.88 97.00 98.37
233122 16.84 22.26 17.20 19.65 20.90 24.53 16.77 18.71 24.35 23.62 23.43 24.10 17.52 23.98 27.70
233123 51.51 51.74 50.86 53.16 56.11 53.58 49.65 49.88 51.94 51.22 60.98 61.67 51.13 55.12 65.29
233124 65.08 66.82 65.57 66.56 66.15 57.71 65.26 64.97 58.04 67.24 70.03 68.84 66.86 67.00 71.67
233126 90.64 58.17 90.42 92.94 92.60 93.99 90.45 88.55 93.98 93.58 93.86 93.87 92.97 94.10 94.02
233130 87.76 87.81 88.98 88.99 70.72 87.99 50.00 85.25 88.35 92.43 50.00 95.81 94.92 91.19 92.53
233131 70.94 69.28 71.59 70.94 71.31 72.14 71.59 71.59 72.32 70.94 72.69 72.42 70.76 70.76 74.26
233132 96.93 98.62 97.52 97.14 94.58 96.85 97.00 97.27 96.90 98.66 98.14 99.15 96.81 96.73 99.05
233133 63.71 65.11 65.00 66.05 64.57 66.21 62.82 64.33 65.98 68.75 66.58 66.28 64.36 64.81 69.18
233134 78.05 75.87 79.05 78.22 80.38 78.38 76.88 78.38 78.38 76.88 77.38 76.54 76.88 76.21 79.22
233135 93.07 92.49 92.10 93.17 93.17 92.10 93.17 93.27 92.10 92.58 92.68 94.53 93.75 93.36 94.05
233137 91.91 93.71 92.16 92.56 90.38 91.58 91.36 88.09 91.60 92.72 92.48 92.39 92.95 92.72 94.01
233142 92.33 96.70 92.90 92.35 63.59 95.47 91.43 93.60 95.56 93.47 93.81 93.25 92.60 93.85 97.17
233143 50.00 92.30 92.76 50.00 70.81 90.28 50.00 50.31 89.26 50.00 50.00 50.00 92.26 50.00 92.53
233146 63.12 60.06 62.79 64.16 63.39 64.42 63.52 54.64 64.21 64.26 64.05 64.57 64.41 64.37 64.28

Table 11: Detailed Table of Results. The test set performance for the plain network, individual
regularization methods and for the regularization cocktails.

21
 Task Id Fixed LR Cocktail Dynamic LR Cocktail
233090 89.270 90.000
233091 99.850 99.850
233092 61.461 56.518
233093 98.000 98.250
233088 74.643 64.881
233094 82.576 79.654
233096 74.381 70.058
233099 82.443 82.551
233102 63.923 63.884
233103 86.619 87.854
233104 99.953 99.967
233106 68.107 69.081
233107 96.826 96.446
233108 95.833 95.370
233109 67.617 67.836
233110 85.993 86.596
233112 80.073 78.985
233113 99.948 83.263
233114 73.546 73.276
233115 87.088 88.077
233116 99.587 99.690
233117 93.723 93.939
233118 91.950 91.964
233119 97.471 98.039
233120 52.668 52.204
233121 98.370 98.522
233122 27.701 28.008
233123 65.287 63.293
233124 71.667 72.243
233126 94.015 93.930
233130 92.535 94.894
233131 74.262 72.140
233132 99.049 99.404
233133 69.183 68.877
233134 79.217 78.887
233135 94.045 94.435
233137 94.010 93.961
233142 97.175 97.106
233143 92.531 92.592
233146 64.280 64.362

Table 12: The test set performances of the regularization cocktails with a fixed initial learning rate
value and a dynamic learning rate chosen by BOHB.

22
 Task Id MLP MLP + D MLP + S XGB. + ES XGB. + ES + ENC XGB. + ENC CatBoost TabN. + ES AutoGL. + HPO MLP + C
233090 84.131 86.916 87.354 90.000 89.000 89.000 90.000 66.678 78.854 89.270
233091 99.701 99.850 99.687 99.701 99.687 99.701 99.197 99.687 99.238 99.850
233092 37.991 38.704 51.321 50.631 48.841 48.779 51.882 30.447 51.943 61.461
233093 97.750 98.000 97.000 96.750 98.000 98.250 97.250 97.250 97.500 98.000
233088 69.405 68.095 66.429 66.786 65.595 68.214 67.976 63.571 77.738 74.643
233094 83.766 82.603 85.444 75.081 73.241 74.405 79.732 66.061 82.624 82.576
233096 70.274 72.980 79.429 50.000 62.713 65.027 65.660 60.065 72.219 74.381
233099 76.893 78.520 81.354 78.790 78.917 79.529 81.045 76.715 82.768 82.443
233102 60.997 63.754 62.709 60.312 61.734 60.703 61.001 54.327 61.119 63.923
233103 87.514 88.387 87.886 87.079 87.841 87.553 87.579 79.027 87.975 86.619
233104 99.971 99.962 99.962 99.962 99.977 99.967 99.968 99.958 99.971 99.953
233106 62.831 67.035 63.560 98.668 98.850 98.721 69.889 53.774 56.662 68.107
233107 95.917 96.232 96.273 96.460 96.785 97.263 97.049 96.232 96.212 96.826
233108 87.500 90.741 93.519 95.370 93.519 93.519 95.370 85.648 94.907 95.833
233109 67.836 68.421 69.956 57.237 57.456 62.427 59.576 64.547 69.152 67.617
233110 78.076 83.145 86.137 73.329 72.087 72.252 74.152 67.893 83.097 85.993
233112 73.627 76.345 75.809 72.849 72.875 73.730 71.578 60.440 76.312 80.073
233113 99.475 99.892 99.494 97.143 98.571 98.571 98.571 63.273 83.851 99.948
233114 67.752 66.873 67.985 72.781 72.779 72.800 72.809 72.431 73.167 73.546
233115 86.268 86.268 83.159 91.186 50.000 89.376 91.016 87.428 86.300 87.088
233116 97.442 99.690 97.757 96.085 93.512 97.870 100.000 88.195 97.565 99.587
233117 94.805 94.589 93.723 93.723 91.991 93.290 92.424 91.126 91.126 93.723
233118 90.464 90.507 89.007 91.064 91.136 91.093 90.893 89.171 90.964 91.950
233119 97.061 97.237 90.997 90.938 90.085 90.748 82.150 88.403 99.270 97.471
233120 50.262 50.301 51.917 51.708 52.257 52.190 52.083 51.340 52.452 52.668
233121 96.125 97.000 2.174 91.967 93.082 92.533 95.777 NaN 97.299 98.370
233122 16.836 23.983 18.618 19.004 22.969 23.190 23.213 20.945 26.466 27.701
233123 51.505 55.118 56.543 54.858 54.705 55.295 55.453 55.212 60.209 65.287
233124 65.081 66.996 63.064 61.342 63.472 64.717 62.660 63.458 67.986 71.667
233126 90.639 94.099 93.701 93.880 94.055 93.979 93.932 49.790 95.172 94.015
233130 87.759 91.194 93.720 89.546 87.304 86.607 87.910 90.012 91.776 92.535
233131 70.941 70.756 69.926 74.815 74.170 75.738 73.708 69.649 73.801 74.262
233132 96.930 96.733 96.462 97.107 96.467 96.519 99.259 97.266 98.856 99.049
233133 63.707 64.814 64.877 68.952 68.911 70.331 71.708 63.811 66.074 69.183
233134 78.048 76.211 75.539 77.867 77.197 78.370 80.052 75.527 80.211 79.217
233135 93.070 93.363 95.217 95.119 94.827 94.924 95.119 92.681 93.364 94.045
233137 91.905 92.724 91.513 89.141 90.793 89.026 NaN 91.334 93.551 94.010
233142 92.331 93.852 92.707 95.019 93.870 94.047 85.379 89.384 96.207 97.175
233143 50.000 50.000 50.000 88.263 87.749 89.791 92.345 85.701 92.829 92.531
233146 63.125 64.367 50.000 57.866 58.143 58.324 56.563 50.340 59.203 64.280

Table 13: The results for the remaining baselines used in our experiments. Each performance
represents the test accuracy of the incumbent configuration after being refit.

23