Scientia Iranica A (2020) 27(6), 2645{2656

Sharif University of Technology

Scientia Iranica
Transactions A: Civil Engineering
http://scientiairanica.sharif.edu

Invited/Review Article

Machine learning in structural engineering

J.P. Amezquita-Sancheza; , M. Valtierra-Rodrigueza , and H. Adelib
a. ENAP-RG, CA Sistemas Dinamicos, Faculty of Engineering, Departments of Electromechanical, and Biomedical Engineering,
Autonomous University of Queretaro, Campus San Juan del Rio, Moctezuma 249, Col. San Cayetano, 76807, San Juan del Rio,
Queretaro, Mexico.
b. Department of Civil, Environmental, and Geodetic Engineering, The Ohio State University, 470 Hitchcock Hall, 2070 Neil
Avenude, Columbus, OH 43220, USA.
Received 14 November 2020; accepted 18 November 2020

KEYWORDS Abstract. This article presents a review of selected articles about structural engineering
Civil structures; applications of Machine Learning (ML) in the past few years. It is divided into the following
Machine learning; areas: structural system identi cation, structural health monitoring, structural vibration
Deep learning; control, structural design, and prediction applications. Deep neural network algorithms
Structural have been the subject of a large number of articles in civil and structural engineering. There
engineering; are, however, other ML algorithms with great potential in civil and structural engineering
System identi cation; that are worth exploring. Four novel supervised ML algorithms developed recently by the
Structural health senior author and his associates with potential applications in civil/structural engineering
monitoring; are reviewed in this paper. They are the Enhanced Probabilistic Neural Network (EPNN),
Vibration control; the Neural Dynamic Classi cation (NDC) algorithm, the Finite Element Machine (FEMa),
Structural design; and the Dynamic Ensemble Learning (DEL) algorithm.
Prediction. © 2020 Sharif University of Technology. All rights reserved.

1. Introduction In general, an ML system consists of three

components: inputs comprising a dataset of sig-
Machine Learning (ML) is a key Arti cial Intelligence nals/images/features, the ML algorithm, and output
(AI) technology that is impacting almost every eld which is associated with the phenomenon studied (see
in a signi cant way from image recognition, e.g., pupil Figure 1). ML algorithms can be classi ed into three
detection [1], multi-object tracking [2], video surveil- broad categories:
lance [3], multi-target regression [4], thermal infrared
face identi cation [5], and human activity recogni- a) Supervised learning such as Support Vector Ma-
tion [6], to various brain and neuroscience applications, chine (SVM) [14], various neural network models,
e.g., building functional brain network [7], motor im- statistical regression, Random Forest (RF) [15],
agery brain-computer interface [8,9], mapping scalp to fuzzy classi ers [16], and Decision Trees (DTs);
intracranial EEG [10], seizure detection [11], diagnosis
of the Parkinson's disease [12], and characterization of b) Unsupervised learning such as various clustering al-
the modulation of the hippocampal rhythms [13]. gorithms, e.g., k-means clustering and hierarchical
clustering [17], autoencoders, self-organizing maps,
competitive learning [18], and deep Boltzmann
*. Corresponding author. Tel.: +52 (1) 427 274 12 44 machine;
E-mail address: [email protected] (J.P.
Amezquita-Sanchez) c) Reinforcement learning such as Q-learning, R-
learning, and Temporal Di erence (TD) learn-
doi :10.24200/sci.2020.22091 ing [19].
2646 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656

Figure 1. Components of an ML system.

The rst journal article on civil engineering appli- associated with the structural behavior. Amezquita-
cation of neural networks was published in 1989 [20]. Sanchez et al. [25] presented a robust methodology for
Amezquita-Sanchez et al. [21] presented a review of identi cation of modal parameters of large smart struc-
research articles on neural networks in civil engineering tures based on the adroit integration of the multiple
published from 2001 to 2016. This article presents signal classi cation algorithm, the empirical wavelet
a review of selected articles on structural engineering transform, and the Hilbert transform [26] and applied it
applications of ML in recent years since 2017. It is to calculate the natural frequencies and damping ratios
divided into the following topics where most of the of a 123-story super high-rise building structure, the
ML research in structural engineering is published: Lotte World Tower, the tallest building in Korea (see
structural system identi cation, structural health mon- Figure 2), subjected to ambient vibrations. The results
itoring, structural vibration control, structural design, showed that the proposed approach could identify the
and prediction applications. In addition, four novel natural frequencies and damping ratios of large civil
supervised ML algorithms developed recently by the structures with high accuracy.
senior author and his associates with potential appli- For a more robust SSI strategy capable of deal-
cations in civil/structural engineering are introduced. ing with the inherent noise, nonlinearities, and un-
certainties present in the acquired samples, Perez-
2. Structural system identi cation Ramirez et al. [27] combined the empirical mode de-
Structural System Identi cation (SSI) is an important
topic in structural engineering as it allows constructing
a mathematical model of a structural system from a set
of input-output measurements generated by dynamic
time series signals [22]. Perez-Ramirez et al. [23]
presented a methodology for identi cation of modal
parameters of structures using ambient vibrations and
Synchrosqueezed Wavelet Transform (SWT). Jiang et
al. [24] introduced a fuzzy stochastic neural network
model for nonparametric identi cation of civil struc-
tures using the nonlinear autoregressive moving aver-
age with exogenous inputs model through the combina-
tion of two computational intelligence techniques, i.e.,
fuzzy logic and neural networks. The proposed model
was validated using a 1:20 scaled model of a 38-storey
concrete building and a benchmark 4-story 2  2 bay
3D steel frame.
Denoising a signal for an e ective SSI scheme
can represent a challenging task because this process
can also inadvertently remove frequency components Figure 2. Lotte World Tower in Seoul, Korea.
 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656 2647

composition [28{30], a recurrent neural network model, with the goal of determining the health condition of the
Bayesian training [31], and mutual information for structure in an automated manner.
response prediction of civil structures subjected to Kostic and Gul [36] combined an autoregressive
extreme loadings. The e ectiveness of the proposed model with exogenous inputs with a Multi-Layer Per-
approach was validated using the experimental data ceptron Neural Network (MLPNN) for damage detec-
of a 1:20-scaled 38-story high-rise building structure tion of a simulated footbridge structure at varying
subjected to arti cial seismic excitations and ambient temperatures. Pan et al. [37] evaluated three time-
vibrations and a ve-story steel frame subjected to frequency methods, Wavelet Transform (WT), Hilbert-
di erent levels of the Kobe earthquake. Huang Transform (HHT), and Teager-Huang Trans-
Yao et al. [32] reported blind modal identi cation form (THT) for identifying features in measured signals
using limited sensors and a modi ed sparse component in combination with SVM using a simulated cable-
analysis. Yuen and Huang [33] introduced a Bayesian stayed bridge.
Frequency-domain substructure Identi cation. Yuen Incipient or light damage represents a challenge
et al. [34] described self-calibrating Bayesian real-time for identi cation. Yanez-Borjas et al. [38] proposed
system identi cation. Tian et al. [35] discussed system the fusion of statistical indices and a decision tree for
identi cation of pedestrian bridges using particle detecting damage due to corrosion in a 3D 9-bay and
image velocimetry. 169-member truss-type bridge subjected to dynamic
excitations. The authors reported that the proposal
could identify light damage due to external corrosion,
3. Structural Health Monitoring (SHM) causing 1 mm reduction in the bar element diame-
Structural Health Monitoring (SHM) continues to ter. Amezquita-Sanchez [39] integrated the Shannon
be the subject of intensive research in structural entropy index with a decision tree for evaluating the
engineering. It can be divided into two categories measured responses of a 1:20 scaled model of a 38-
of image-based SHM employing the computer vision storey concrete building structure under di erent levels
technology and vibration signal-based SHM based of damage produced by cracks.
on the signals obtained during dynamic events. The aforementioned works have exhibited ad-
The latter in turn can be divided into two general vances in SHM; however, they require a hand-crafted
approaches: parametric system identi cation (modal feature extraction approach to e ective classi cation in
parameters identi cation) and non-parametric the subsequent step [40]. In recent years, deep learning
system identi cation. ML algorithms have been used algorithms such as Convolutional Neural Networks
extensively in both types of SHM. (CNNs) have been employed for automatic feature ex-
traction in SHM. In these methods, feature extraction
3.1. Vibration signal-based SHM and classi cation steps are performed in a single step
SHM based on the non-parametric system identi - to avoid the exhaustive tests between features and
cation approach consists of two main stages of fea- classi ers (see Figure 3) [41,42]. Abdeljaber et al. [43]
ture extraction/selection and classi cation. The fea- explored a 1D-CNN for determining the condition of a
ture/patterns identi ed in the rst step are employed benchmark 4-story 2  2 bay 3D steel frame subjected
for designing and training a machine learning algorithm to ambient vibrations. Krishnasamy and Arumulla [44]

Figure 3. (a) Traditional machine learning and (b) deep learning.

2648 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656

combined a second-order blind identi cation with WT al. [55] described a context-aware deep convolutional
and autoregressive time series models for detecting semantic segmentation network for detecting cracks in
minor incipient damage such as subtle cracks in a beam structures. Wu et al. [56] discussed pruning CNNs
subjected to dynamic excitations. for ecient edge computing in health condition as-
The e ective training of supervised ML ap- sessment of structures. Nayyeri et al. [57] described a
proaches requires a large set of data from healthy foreground-background separation technique for bridge
and damaged structures. To overcome this limitation, crack detection.
unsupervised ML-based methods have been proposed Deng et al. [58] employed CNN for concrete crack
recently because they do not require labeling the train- detection with handwriting script interferences. Pan
ing data from di erent damage scenarios. Ra ei and and Yang [59] described a post-disaster image-based
Adeli [45] presented a novel unsupervised deep learning damage detection of reinforced concrete buildings using
model for global and local health condition assessment dual CNNs. Liu et al. [60] reported an image-based
of structures using ambient vibration response through crack assessment of bridge piers employing Unmanned
integration of SWT, fast Fourier transform, and deep Aerial Vehicles (UAVs) and 3D scene reconstruction.
restricted Boltzmann machine. The model extracts Jiang and Zhang [61] also discussed a real-time crack
features from the frequency domain of the recorded assessment using deep neural networks and wall-
signals automatically. climbing UAVs.
Ibrahim et al. [46] compared the classi cation per- Athanasiou et al. [62] outlined a machine learning
formance of three machine learning algorithms, SVM, approach to crack assessment of reinforced concrete
K-Nearest Neighbor (KNN), and CNN, for evaluating shells using multifractal analysis as a feature extractor.
the health condition of two simulated four- and eight-
story building structures subjected to earthquakes. 4. Vibration control of structures
They reported that CNN outperformed SVM and KNN
in terms of accuracy for damage detection. Zhang Dynamic loadings such as trac, wind, and seismic
et al. [47] discussed vibration-based structural state activity generate vibrational responses that can neg-
identi cation by a one-dimensional CNN. Huang et atively a ect the integrity of a structure. In order to
al. [48] presented a multitask sparse Bayesian learning decrease this negative impact, many e orts on research
for SHM applications. and technology development for structural vibration
Wang et al. [49] described shear loading detection control have been proposed.
of through bolts in bridges using a percussion-based Figure 4 presents the four classes of vibration
one-dimensional memory-augmented CNN. Naranjo- control technologies and a tabular summary of sys-
Perez et al. [50] presented a collaborative machine tems proposed in each category. Ying and Ni [63]
learning-optimization algorithm to improve the nite reviewed applications of the magnetorheological visco-
element model updating of structures. Their proposal elastomer materials in structural vibration control. Lu
consists of the harmonic search and active-set algo- et al. [64] reviewed various non-linear dampers. Deng
rithms, multilayer perceptron neural networks, and the and Dapino [65] reviewed magnetostrictive materials
principal component analysis, where advantages such for structural vibration control such as Terfenol-D and
as the computation time, robustness and e ectiveness Galfenol (alloy of iron and gallium). Elias and Mat-
of an actual steel footbridge model are obtained. From sagar [66] reviewed vibration control of structures using
this work, it is observed that the combination of several passive Tuned Mass Dampers (TMD) for wind- and
machine-learning algorithms and other mathematical earthquake-excited structures. Rahimi et al. [67] also
tools can lead to more powerful solution methods. reviewed the application of TMDs for vibration control
Wang and Cha [51] combined a deep auto- of structures including the theoretical backgrounds for
encoder, an unsupervised deep learning method, with a various types of TMDs and practical and economic
one-class SVM for vibration-based health monitoring of aspects.
a laboratory-scaled steel bridge. The authors reported Computational intelligence approaches such as
an accuracy rate of 91% for light damage detection. neural networks, fuzzy logic systems [68], and ge-
Sajedi and Liang [52] discussed the vibration-based netic algorithms [69] and their combination have
semantic damage segmentation for SHM. played a signi cant role in the development of
adaptive/intelligent control algorithms. Wang and
3.2. Image-based SHM Adeli [70] presented a self-constructing Wavelet Neural
Cha et al. [53] presented the autonomous structural Network (WNN) for nonlinear adaptive control of
visual inspection using region-based deep learning structures based on integration of a self-constructing
for detecting di erent types of damage. Gao and wavelet neural network developed speci cally for struc-
Mosalam [54] employed a deep transfer learning for tural system identi cation with an adaptive fuzzy
image-based structural damage recognition. Zhang et sliding mode control approach. The authors note
 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656 2649

Figure 4. Classi cation of di erent vibration control systems.

\the algorithm is particularly suitable when the physical ferent dynamic systems and fuzzy logic allows dealing
properties such as the sti nesses and damping ratios of with the information uncertainty usually existing in
the structural system are unknown or partially known real-world problems.
which is the case when a structure is subjected to Rahmani et al. [75] employed a reinforcement
an extreme dynamic event such as an earthquake as learning, Q-learning, method for controlling the vi-
the structural properties change during the event." A bration of a moment frame subjected to seismic load-
unique aspect of the algorithm is that the structural ing. Lu et al. [76] proposed a vibration identi cation
identi cation and control are performed simultane- method based on a CNN to nd the optimal parameters
ously, which makes it adaptive and more suitable for of Linear Quadratic Regulator (LQR) algorithm for
real-life structures. A fuzzy compensation controller vibration control of a single-degree-of-freedom linear
reduces the chattering and adaptive laws based on Lya- system subjected to dynamic excitations. The authors
punov functions for nding the unknown parameters of note that the integration of CNN with LQR improves
the WNN provide control stability. the performance of the LQR algorithm with xed
An adaptive WNN controller was introduced by parameters; however, it requires collecting a large
Chen and Liang [71] to improve the performance of a amount of vibration input data and its ecacy of real
pneumatic isolation system. Bui et al. [72] discussed a structures needs to be investigated.
fuzzy controller for structural vibration control. Wang
et al. [73] described a fuzzy nite-time stable compen- 5. Structural design
sation control considering di erent actuator failures.
Fu et al. [74] described a fuzzy-neural network control Application of machine learning in engineering design
for a magnetorheological elastomer vibration isolation was pioneered by Adeli and Yeh [77] three decades
system. ago. Ra ei et al. [78] presented a novel approach
From the reviewed works, it is observed that to concrete mixed design using the patented neural
neural network- and fuzzy logic-based approaches are dynamics optimization model of Adeli and Park [79]
the most common solutions for vibration control; yet, and a classi cation algorithm used as a virtual lab to
their combination yields more powerful results since the predict whether desired constraints are satis ed in each
advantages of both computing approaches are exploited iteration of the design or not. The authors tested the
synergistically. For instance, neural networks provide model using three di erent classi cation algorithms:
the capability to deal with non-linear properties of dif- SVM, Probabilistic Neural Network (PNN), and En-
2650 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656

hanced Probabilistic Neural Network (EPNN) to be 7. New supervised learning algorithms

described later. They reported EPNN with the highest
accuracy. Deep neural network algorithms such as CNN have
Zheng et al. [80] proposed using neural network been the subject of a large number of articles in
learning to accelerate the time-consuming 3D form civil and structural engineering so much so that their
nding and topological design of compression-only shell applications are becoming almost routine. There are,
structures with planar faces. however, other new and powerful supervised learning
or classi cation algorithms with great potential in civil
6. Prediction applications and structural engineering that are worth exploring.
Four of them developed by the senior author and his
Greco et al. [81] used a genetic algorithm to predict associates are reviewed in this section.
seismic collapse of frame structures. Asteris and
Nikoo [82] employed an arti cial bee colony-based 7.1. Enhanced probabilistic neural network
neural network to predict the fundamental period of Ahmadlou and Adeli [92] proposed a supervised clas-
in lled frame structures using a number of stories and si er called Enhanced Probabilistic Neural Network
spans, the span length, the in ll wall panel sti ness, (EPNN) to improve the accuracy and robustness of
and the percentage of openings as the input of the PNN by means of local decision circles with the purpose
neural network model. Luo and Paal [83] described of incorporating the non-homogeneity often existing in
a locally weighted SVM model to predict the drift a training population. They demonstrated that EPNN
capacity of reinforced concrete columns for seismic was superior to PNN by applying it to three di erent
vulnerability assessments. benchmark classi cation problems: iris data, diabetic
Deep learning approaches have also been reported data, and breast cancer data. Since then, EPNN
for prediction tasks. Nie et al. [84] used a CNN has been demonstrated to be a powerful classi cation
for stress eld prediction in cantilevered structures. algorithm for diagnosis of various neurological disorders
Nguyen et al. [85] employed a deep neural network with such as Parkinson's disease [93] and Mild Cognitive Im-
high-order neuron to predict the compressive strength pairment (MCI) [94] and prediction of sudden cardiac
of foamed concrete where a cross-entropy cost function arrest [95]. To the best of the authors' knowledge, there
is used to enhance the model performance. Luo is only one structural engineering application of EPNN,
and Kareem [86] presented deep convolutional neural as described in Section 5.
networks for uncertainty propagation in random elds
and statistical characterization of the system responses
under various spatially varying properties. 7.2. Neural dynamic classi cation algorithm
Oh et al. [87] employed a CNN to estimate Starting with the patented neural dynamics optimiza-
the response of tall buildings due to wind loading. tion model of Adeli and Park [79], Ra ei and Adeli [45]
Similarly, Oh et al. [88] employed a CNN to predict the introduced a new supervised classi cation algorithm,
seismic responses of building structures. The viability called Neural Dynamic Classi cation (NDC), with the
of the CNN was validated using numerical simulation goal of uncovering the most e ective feature spaces
and experimental data obtained on a 3-story concrete and nding the optimum number of features required
structure subjected to seismic excitations. Zhang et for accurate classi cation. The algorithm is capable
al. [89] evaluated a deep learning approach called of solving highly complicated classi cation problems
Long Short-Term Memory (LSTM) network [90] for by employing a new feature space with large margins
nonlinear structural response prediction of a 6-story between clusters and close proximity of the classmates
hotel building located at San Bernardino, California, and a set of transformation functions. They compared
subjected to seismic excitations. In addition, an the new classi cation algorithm with the Probabilis-
unsupervised learning algorithm, K-means clustering, tic Neural Network (PNN), EPNN, and SVM using
was employed to cluster the seismic inputs for: multiple standard benchmark problems. They note
\NDC yields the most accurate classi cation results
1. Generating the most informative datasets for train- followed by EPNN. A beauty of the new algorithm
ing the LSTM, is the smoothness of convergence curves which is an
2. Improving the prediction accuracy with limited indication of robustness and good performance of the
data. algorithm."
So far, NDC has been used successfully to solve
They reported an accuracy rate of 90% for predict- two complicated structural engineering problems, that
ing the structural response. Gulgec et al. [91] also is, development of an earthquake early warning sys-
discussed the application of deep learning to strain tem [96] and damage detection in high-rise building
estimation from acceleration data for fatigue. structures [97].
 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656 2651

7.3. Finite element machine for fast learning novel supervised ML algorithms developed recently by
Recently, Pereira et al. [98] proposed a new super- the senior author and his associates with potential
vised pattern classi er, called Finite Element Machine applications in civil/structural engineering were intro-
(FEMa), with its research ideology and theoretical duced in this paper.
basis in the Finite Element Method widely used in nu-
merical analysis and solution of numerous engineering References
problems. In this model, \each training sample is the
center of a basis function, and the whole training set 1. Vera-Olmos, F.J., Pardo, E., Melero, H., and
is modeled as a probabilistic manifold for classi cation Malpica, N. \DeepEye: Deep convolutional net-
purpose." The algorithm is parameterless and does work for pupil detection in real environments", Inte-
not require a training step, which can be a great grated Computer-Aided Engineering, 26(1), pp. 85{95
advantage when the dataset is large, that is so-called (2019).
the Big Data problems. FEMa has been shown to 2. Yang, T., Cappelle, C., Ruichek, Y., and El Bagdouri,
yield competitive results for classifying 23 di erent M. \Multi-object tracking with discriminant corre-
public benchmark datasets (e.g., breast data, wine lation lter based deep learning tracker", Integrated
data, diabetes data, among others) compared with nine Computer-Aided Engineering, 26(3), pp. 273{284
other pattern classi ers such as DT, Bayesian [99], (2019).
KNN, RF, Optimum-Path Forest (OPF), SVM with 3. Benito-Picazo, J., Domnguez, E., Palomo, E.J.,
Radial Basis Function, SVM with a sigmoid function, and Lopez-Rubio, E. \Deep learning-based video
and EPNN. The authors believe that FEMa has great surveillance system managed by low cost hardware
potentials and should be explored for civil/structural and panoramic cameras", Integrated Computer-Aided
Engineering, 27(4), pp. 373{387 (2020).
engineering applications.
4. Reyes, O. and Ventura, S. \Performing multi-target
7.4. Dynamic ensemble learning algorithm regression via a parameter sharing-based deep net-
As noted in this article, many di erent ML algorithms work", International Journal of Neural Systems,
have been developed over the past three decades. 29(9), 1950014 (22 pages) (2019).
Ensemble ML methods employ multiple learning al- 5. Wang, P. and Bai, X. \Regional parallel structure
gorithms to achieve more accurate results than those based CNN for thermal infrared face identi cation",
obtained using any of the constituent learning algo- Integrated Computer-Aided Engineering, 25(3), pp.
rithms [100{102]. Rokibul Alam et al. [103] presented 247{260 (2018).
a Dynamic Ensemble Learning (DEL) algorithm for 6. Rodriguez-Lera, F.J., Rico, F.M., and Olivera, V.M.
designing an ensemble of neural networks. According \Neural networks for recognizing human activities in
to the authors, \DEL algorithm determines the size home-like environments", Integrated Computer-Aided
of ensemble, the number of individual NNs employing Engineering, 26(1), pp. 37{47 (2019).
a constructive strategy, the number of hidden nodes 7. Hua, C., Wang, H., Wang, H., Lu, S., Liu, C., and
of individual NNs employing a constructive-pruning Khalid, S.M. \A novel method of building functional
strategy, and di erent training samples for individual brain network using deep learning algorithm with
NN's learning." They introduced the concept of application in pro ciency detection", International
negative correlation learning to enhance the diversity Journal of Neural Systems, 29(1), 1850015 (17 pages)
in learning. The ecacy of the model has been veri ed (2019).
by application to four medical and seven non-medical 8. Gaur, P., McCreadie, K., Pachori, R.B., Wang,
benchmark problems and datasets. H., and Prasad, G. \Tangent space feature-based
transfer learning classi cation model for two-class
motor imagery brain-computer interface", Interna-
8. Conclusions tional Journal of Neural Systems, 29(10), 19500215
(17 pages) (2019).
There has been a signi cant interest in the application 9. Lozano, A., Suarez, J.S., Soto-Sanchez, C., Garrigos,
of ML algorithms in structural engineering in recent J., Martinez-Alvarez, J.J., Ferrandez, J.M., and
years. Some of the recent structural engineering Fernandez, E. \Neurolight: a deep learning neural
applications of ML algorithms were reviewed in this interface for cortical visual prostheses", International
paper. The most common ML approaches used in Journal of Neural Systems, 30(9), 2050045 (19 pages)
structural engineering included SVM and deep learning (2020).
algorithms such as CNN and LSTM. The latter has 10. Antoniades, A., Spyrou, L., Martin-Lopez, D.,
captured the imagination of structural engineering Valentin, A., Alarcon, G., Sanei, S., and Took,
researchers in the past four years. There are, however, C.C. \Deep neural architectures for mapping scalp to
other ML algorithms with great potential in civil and intracranial EEG", International Journal of Neural
structural engineering that are worth exploring. Four Systems, 28(8), 1850009 (2018).
2652 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656

11. Ansari, A.H., Cherian, P.J., Caicedo, A., Naulaers, selection and parametric identi cation", Computer-
G., De Vos, M., and Van Hu el, S. \Neonatal Aided Civil and Infrastructure Engineering, 30(10),
seizure detection using deep convolutional neural pp. 785{801 (2015).
networks", International Journal of Neural Systems, 23. Perez-Ramirez, C.A., Amezquita-Sanchez, J.P.,
29(4), 1850011 (20 pages) (2019). Adeli, H. Valtierra-Rodriguez, M., Camarena-
12. Manzanera, M.O., Meles, S.K., Leenders, K.L., Martinez, D., and Rene Romero-Troncoso, R.J. \New
Renken, R.J., Pagani, M., Arnaldi, D., Nobili, F., methodology for modal parameters identi cation of
Obeso, J., Oroz, M.R., Morbelli, S., and Maurits, smart civil structures using ambient vibrations and
N.M. \Scaled subpro le modeling and convolutional synchrosqueezed wavelet", Engineering Applications
neural networks for the identi cation of Parkinson's of Arti cial Intelligence, 48, pp. 1{16 (2016).
disease in 3D nuclear imaging data", International
Journal of Neural Systems, 29(9), 1950010 (15 pages) 24. Jiang, X., Mahadevan, S., and Yuan, Y. \Fuzzy
(2019). stochastic neural network model for structural sys-
tem identi cation", Mechanical Systems and Signal
13. Park, S.E., Laxpati, N.G., Gutekunst, C.A., Con- Processing, 82, pp. 394{411 (2017).
nolly, M.J., Tung, J., Berglund, K., Mahmoudi,
B., and Gross, R.E. \A machine learning approach 25. Amezquita-Sanchez, J.P., Park, H.S., and Adeli, H.
to characterize the modulation of the hippocampal \A novel methodology for modal parameters iden-
rhythms via optogenetic stimulation of the medial ti cation of large smart structures using MUSIC,
septum", International Journal of Neural Systems, empirical wavelet transform, and Hilbert transform",
29(10), 1950020 (21 pages) (2019). Engineering Structures, 147, pp. 148{159 (2017).
14. Zhang, J., Xiao, M., Gao, L., and Chu, S. \Proba- 26. Karami, K., Fatehi, P., and Yazdani, A. \On-line
bility and interval hybrid reliability analysis based on system identi cation of structures using wavelet-
adaptive local approximation of projection outlines Hilbert transform and sparse component analysis",
using support vector machine", Computer-Aided Civil Computer-Aided Civil and Infrastructure Engineer-
and Infrastructure Engineering, 34(11), pp. 991{1009 ing, 35(8), pp. 870{886 (2020).
(2019). 27. Perez-Ramirez, C.A., Amezquita-Sanchez, J.P.,
15. Yu, B., Wang, H., Shan, W., and Yao, B. \Prediction Valtierra-Rodriguez, M., and Adeli, H. \Recurrent
of bus travel time using random forests based on near neural network model with Bayesian training and
neighbors", Computer-Aided Civil and Infrastructure mutual information for response prediction of large
Engineering, 33(4), pp. 333{350 (2018). buildings", Engineering Structures, 178, pp. 603{615
16. Abbasi, H., Bennet, L., Gunn, A.J., and Unsworth, (2019).
C.P. \Latent phase detection of hypoxic-ischemic 28. Wu, W.H., Chen, C.C., Jhou, J.W., and Lai, G.
spike transients in the EEG of preterm fetal sheep us- \A rapidly convergent empirical mode decomposition
ing reverse biorthogonal wavelets & fuzzy classi er", method for analyzing the environmental temperature
International Journal of Neural Systems, 29(10), e ects on stay cable force", Computer-Aided Civil
1950013, (17 pages) (2019). and Infrastructure Engineering, 33(8), pp. 672{690
17. Chen, Z. and Liu, X.C. \Roadway asset inspec- (2018).
tion sampling using high-dimensional clustering and 29. Zhang, S., Zhou, L., Chen, X., Zhang, L., Li, L.,
locality-sensitivity hashing", Computer-Aided Civil and Li, M. \Network-wide trac speed forecasting:
and Infrastructure Engineering, 34(2), pp. 116{129 3D convolutional neural network with ensemble em-
(2019). pirical mode decomposition", Computer-Aided Civil
18. Lopez-Rubio, E., Molina-Cabello, M.A., Luque- and Infrastructure Engineering, 35(10), pp. 1132{
Baena, R.M., and Dominguez, E. \Foreground de- 1147 (2020).
tection by competitive learning for varying input dis- 30. Saadaoui, F. and Ben Messaoud, O. \Multiscaled
tributions", International Journal of Neural Systems, neural autoregressive distributed lag: a new empir-
28(5), 1750056 (16 pages) (2018). ical mode decomposition model for nonlinear time
19. Sutton, R. and Barto, A., Reinforcement Learning, series forecasting", International Journal of Neural
2nd Ed., MIT Press (2018). Systems, 30(8), 2050039 (15 pages) (2020).
20. Adeli, H. and Yeh, C. \Perceptron learning in en- 31. Wang, J., Liu, X.Z., and Ni, Y.Q. \A Bayesian
gineering design", Computer-Aided Civil and Infras- probabilistic approach for acoustic emission-based
tructure Engineering, 4(4), pp. 247{256 (1989). rail condition assessment", Computer-Aided Civil and
21. Amezquita-Sanchez, J.P., Valtierra-Rodriguez, M., Infrastructure Engineering, 33(1), pp. 21{34 (2018).
Aldwaik, M., and Adeli, H. \Neurocomputing in civil 32. Yao, X.J., Yi, T.H., Qu, C., and Li, H.N.
infrastructure", Scientia Iranica, 23(6), pp. 2417{ \Blind modal identi cation using limited sensors
2428 (2016). through modi ed sparse component analysis by time-
22. Yuen, K.V. and Mu, H.Q. \Real-time system identi- frequency method", Computer-Aided Civil and In-
cation: an algorithm for simultaneous model class frastructure Engineering, 33(9), pp. 769{782 (2018).
 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656 2653

33. Yuen, K.X. and Huang, K. \Identi ability-enhanced 45. Ra ei, M.H. and Adeli, H. \A new neural dynamic
Bayesian frequency-domain substructure identi ca- classi cation algorithm", IEEE Transactions on Neu-
tion", Computer-Aided Civil and Infrastructure En- ral Networks and Learning Systems, 28(12), pp.
gineering, 33(9), pp. 800{812 (2018). 3074{3083 (2017).
34. Yuen, K.V., Kuok, S.C., and Dong, L. \Self- 46. Ibrahim, A., Eltawil, A., Na, Y., and El-Tawil, S.
calibrating Bayesian real-time system identi cation", \A machine learning approach for structural health
Computer-Aided Civil and Infrastructure Engineer- monitoring using noisy data sets", IEEE Transactions
ing, 34(9), pp. 806{821 (2019). on Automation Science and Engineering, 17(2), pp.
35. Tian, Y., Zhang, J., and Yu, S. \Rapid impact 900{908 (2019).
testing and system identi cation of footbridges using 47. Zhang, Y., Miyamori, Y., Mikami, S., and Saito,
particle image velocimetry", Computer-Aided Civil T. \Vibration-based structural state identi cation
and Infrastructure Engineering, 34(2), pp. 130{145 by a 1-dimensional convolutional neural network",
(2019). Computer-Aided Civil and Infrastructure Engineer-
ing, 34(9), pp. 822{839 (2019).
36. Kostic, B. and Gul, M. \Vibration-based damage
detection of bridges under varying temperature ef- 48. Huang, Y., Beck, J.L., and Li, H. \Multitask sparse
fects using time-series analysis and arti cial neural Bayesian learning with applications in structural
networks", Journal of Bridge Engineering, 22(10), health monitoring", Computer-Aided Civil and In-
04017065 (2017). frastructure Engineering, 34(9), pp. 732{754 (2019).
37. Pan, H., Azimi, M., Yan, F., and Lin, Z. \Time- 49. Wang, F., Song, G., and Mo, Y.L. \Shear load-
frequency-based data-driven structural diagnosis and ing detection of through bolts in bridge struc-
damage detection for cable-stayed bridges", Journal tures using a percussion-based one-dimensional
of Bridge Engineering, 23(6), 04018033 (2018). memory-augmented convolutional neural network",
Computer-Aided Civil and Infrastructure Engineering
38. Yanez-Borjas, J.J., Valtierra-Rodriguez, M., (In press) (2020). DOI: 10.1111/mice.12602
Camarena-Martinez, D., and Amezquita-Sanchez,
J.P. \Statistical time features for global corrosion 50. Naranjo-Perez, J., Infantes, M., Jimenez-Alonso,
assessment in a truss bridge from vibration signals", J.F., and Saez, A. \A collaborative machine learning-
Measurement, 160, 107858 (2020). optimization algorithm to improve the nite element
model updating of civil engineering structures", En-
39. Amezquita-Sanchez, J.P. \Entropy algorithms for gineering. Structures, 225, 111327 (2020).
detecting incipient damage in high-rise buildings sub- 51. Wang, Z. and Cha, Y.J. \Unsupervised deep learning
jected to dynamic vibrations", Journal of Vibration approach using a deep auto-encoder with a one-class
and Control (In press) (2020). support vector machine to detect structural damage",
DOI: 10.1177/1077546320929145 Structural Health Monitoring (In press) (2020).
40. Ettouney, M.M. and Alampalli, S., Infrastructure DOI: 10.1177/1475921720934051
Health in Civil Engineering: Theory and Components, 52. Sajedi, S.O. and Liang, X. \Vibration-based semantic
CRC press (2016). damage segmentation for large-scale structural health
41. Torres, J.F., Galicia, A., Troncoso, A., and Martnez- monitoring", Computer-Aided Civil and Infrastruc-
 lvarez, F. \A scalable approach based on deep
A ture Engineering, 35(6), pp. 579{596 (2020).
learning for big data time series forecasting", Inte- 53. Cha, Y.J., Choi, W., Suh, G., Mahmoudkhani,
grated Computer-Aided Engineering, 25(4), pp. 335{ S., and Buyukozturk, O. \Autonomous structural
348 (2018). visual inspection using region-based deep learning for
42. Valtierra-Rodriguez, M., Rivera-Guillen, J.R., detecting multiple damage types", Computer-Aided
Basurto-Hurtado, J.A. De-Santiago-Perez, J.J., Civil and Infrastructure Engineering, 33(9), pp. 731{
Granados-Lieberman, D., and Amezquita-Sanchez, 747 (2018).
J.P. \Convolutional neural network and motor 54. Gao, Y. and Mosalam, K.M. \Deep transfer learn-
current signature analysis during the transient state ing for image-based structural damage recognition",
for detection of broken rotor bars in induction Computer-Aided Civil and Infrastructure Engineer-
motors", Sensors, 20(13), 3721 (2020). ing, 33(9), pp. 748{768 (2018).
43. Abdeljaber, O., Avci, O., Kiranyaz, M.S., Boashash, 55. Zhang, X., Rajan, D., and Story, B. \Concrete crack
B., Sodano, H., and Inman, D.J. \1-D CNNs for struc- detection using context-aware deep semantic segmen-
tural damage detection: Veri cation on a structural tation network", Computer-Aided Civil and Infras-
health monitoring benchmark data", Neurocomput- tructure Engineering, 34(11), pp. 951{971 (2019)
ing, 275, pp. 1308{1317 (2018). 56. Wu, R.T., Ankush Singla, A., Jahanshahi, M.R.,
44. Krishnasamy, L. and Arumulla, R.M.R. \Baseline- Bertino, E., Ko, B.J., and Verma, D. \Pruning
free hybrid diagnostic technique for detection of deep convolutional neural networks for ecient edge
minor incipient damage in the structure", Jour- computing in condition assessment of civil infras-
nal of Performance of Constructed Facilities, 33(2), tructures", Computer-Aided Civil and Infrastructure
04019018 (2019). Engineering, 34(9), pp. 774{789 (2019).
2654 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656

57. Nayyeri, F., Hou, L., Zhou, J., and Guan, H. 70. Wang, N. and Adeli, H. \Self-constructing wavelet
\Foreground-background separation technique for neural network algorithm for nonlinear control of
road and bridge crack detection", Computer-Aided large structures", Engineering Applications of Arti-
Civil and Infrastructure Engineering, 34(6), pp. 457{ cial Intelligence, 41, pp. 249{258 (2015).
470 (2019). 71. Chen, H.Y., and Liang, J.W. \Adaptive wavelet neu-
58. Deng, J., Lu, Y., and Lee, V.C.S. \Concrete crack ral network controller for active suppression control
detection with handwriting script interferences using of a diaphragm-type pneumatic vibration isolator",
faster region-based convolutional neural network", International Journal of Control, Automation and
Computer-Aided Civil and Infrastructure Engineer- Systems, 15(3), pp. 1456{1465 (2017).
ing, 35(4), pp. 373{388 (2020).
72. Bui, H.L., Le, T.A., and Bui, V.B. \Explicit formula
59. Pan, X. and Yang, T.Y. \Post-disaster imaged-based of hedge-algebras-based fuzzy controller and applica-
damage detection and repair cost estimation of re- tions in structural vibration control", Applied Soft
inforced concrete buildings using dual convolutional Computing, 60, pp. 150{166 (2017).
neural networks", Computer-Aided Civil and Infras-
tructure Engineering, 35(5), pp. 495{510 (2020). 73. Wang, J., Huang, Y., Wang, T., Zhang, C., and Liu,
60. Liu, Y.F., Nie, X., Fan, J.S., and Liu, X.G. \Image- Y. \Fuzzy nite-time stable compensation control for
based crack assessment of bridge piers using un- a building structural vibration system with actuator
manned aerial vehicles and 3D scene reconstruction", failures", Applied Soft Computing, 93, 106372 (2020).
Computer-Aided Civil and Infrastructure Engineer- 74. Fu, J., Lai, J., Yang, Z., Bai, J., and Yu, M. \Fuzzy-
ing, 35(5), pp. 511{529 (2020). neural network control for a magnetorheological elas-
61. Jiang, S. and Zhang, J. \Real-time crack assessment tomer vibration isolation system", Smart Materials
using deep neural networks with wall-climbing un- and Structures, 29(7), 074001 (2020).
manned aerial system", Computer-Aided Civil and In- 75. Rahmani, H.R., Chase, G., Wiering, M., and Konke,
frastructure Engineering, 35(6), pp. 549{564 (2020). C. \A framework for brain learning-based control of
62. Athanasiou, A., Ebrahimkhanlou, A., Zaborac, J., smart structures", Advanced Engineering Informat-
Hrynyk, T., and Salamone, S. \A machine learning ics, 42, 100986 (2019).
approach based on multifractal features for crack 76. Lu, X., Liao, W., Huang, W., Xu, Y., and
assessment of reinforced concrete shells", Computer- Chen, X. \An improved linear quadratic
Aided Civil and Infrastructure Engineering, 35(6), regulator control method through convolutional
pp. 565{578 (2020). neural network-based vibration identi cation",
63. Ying, Z. and Ni, Y. \Advances in structural vi- Journal of Vibration and Control (In press)
bration control application of magneto-rheological https://doi.org/10.1177/1077546320933756 (2020).
visco-elastomer", Journal of Theoretical and Applied DOI: 10.1177/1077546320933756
Mechanics, 7(2), pp. 61{66 (2017). 77. Adeli, H. and Yeh, C. \Explanation-based machine
64. Lu, Z., Wang, Z., Zhou, Y., and Lu, X. \Nonlinear learning in engineering design", Engineering Appli-
dissipative devices in structural vibration control: A cations of Arti cial Intelligence, 3(2), pp. 127{137
review", Journal of Sound and Vibration, 423, pp. (1990).
18{49 (2018).
78. Ra ei, M.H., Khushefati, W.H., Demirboga, R., and
65. Deng, Z. and Dapino, M.J. \Review of magnetostric- Adeli, H. \Novel approach for concrete mix design
tive materials for structural vibration control", Smart using neural dynamics model and the virtual lab
Materials and Structures, 27(11), 113001 (2018). concept", ACI Materials Journal, 114(1), pp. 117{
66. Elias, S. and Matsagar, V. \Research developments 127 (2017).
in vibration control of structures using passive tuned 79. Adeli, H. and Park, H.S., Neurocomputing for Design
mass dampers", Annual Reviews in Control, 44, pp. Automation, Boca Raton, Florida: CRC Press (1998).
129{156 (2017).
67. Rahimi, F., Aghayari, R., and Samali, B. \Applica- 80. Zheng, H., Moosavi, V., and Akbarzadeh, M. \Ma-
tion of tuned mass dampers for structural vibration chine learning assisted evaluations in structural de-
control: a state-of-the-art review", Civil Engineering sign and construction", Automation in Construction,
Journal, 6(8), pp. 1622{1651 (2020). 119, 103346 (2020).
68. Chiew, F.H., Ng, C.K., Chai, K.C., and Tay, K.M. 81. Greco, A., Cannizzaro, F., and Pluchino, A. \Seismic
\A fuzzy adaptive resonance theory-based model collapse prediction of frame structures by means of
for mix proportion estimation of high performance genetic algorithms", Engineering Structures, 143, pp.
concrete", Computer-Aided Civil and Infrastructure 152{168 (2017).
Engineering, 32(9), pp. 772{786 (2017). 82. Asteris, P.G., and Nikoo, M. \Arti cial bee colony-
69. Padillo, F., Luna, J.M., Herrera, F., and Ventura, S. based neural network for the prediction of the fun-
\Mining association rules on big data through mapre- damental period of in lled frame structures", Neural
duce genetic programming", Integrated Computer- Computing and Applications, 31(9), pp. 4837{4847
Aided Engineering, 25(1), pp. 31{48 (2018). (2019).
 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656 2655

83. Luo, H. and Paal, S.G. \A locally weighted ma- transform-homogeneity model for sudden cardiac
chine learning model for generalized prediction of death prediction using ECG signals", Journal of
drift capacity in seismic vulnerability assessments", Medical Systems, 42(10), p. 176 (2018).
Computer-Aided Civil and Infrastructure Engineer- 96. Ra ei, M.H. and Adeli, H. \NEEWS: A novel earth-
ing, 34(11), pp. 935{950 (2019). quake early warning model using neural dynamic
84. Nie, Z., Jiang, H., and Kara, L.B. \Stress eld classi cation and neural dynamic optimization", Soil
prediction in cantilevered structures using convolu- Dynamics and Earthquake Engineering, 100, pp. 417{
tional neural networks", Journal of Computing and 427 (2017).
Information Science in Engineering, 20(1), pp. 1{16 97. Ra ei, M.H. and Adeli, H. \A novel machine learning-
(2020). based algorithm to detect damage in high-rise build-
85. Nguyen, T., Kashani, A., Ngo, T., and Bordas, S. ing structures", The Structural Design of Tall and
\Deep neural network with high-order neuron for the Special Buildings, 26(18), e1400 (2017).
prediction of foamed concrete strength", Computer- 98. Pereira, D.R., Piteri, M.A., Souza, A.N., Papa, J.P.,
Aided Civil and Infrastructure Engineering, 34(4), and Adeli, H. \FEMa: a nite element machine for
pp. 316{332 (2019). fast learning", Neural Computing and Applications,
86. Luo, X. and Kareem, A. \Deep convolutional neu- 32(10), pp. 6393{6404 (2019).
ral networks for uncertainty propagation in random 99. Schetinin, V., Jakaite, L., and Krzanowski, W.
elds", Computer-Aided Civil and Infrastructure En- \Bayesian learning of models estimating uncertainty
gineering, 34(12), pp. 1041{1054 (2019). in alert systems: application to aircraft collision
87. Oh, B.K., Glisic, B., Kim, Y., and Park, H.S. \Convo- avoidance", Integrated Computer-Aided Engineering,
lutional neural network-based wind-induced response 25(3), pp. 229{245 (2018).
estimation model for tall buildings", Computer-Aided 100. Molina-Cabello, M.A., Luque-Baena, R.M., Lopez-
Civil and Infrastructure Engineering, 34(10), pp. Rubio, E., and Thurnhofer-Hemsi, K. \Vehicle
843{858 (2019). type detection by ensembles of convolutional neu-
88. Oh, B.K., Park, Y., and Park, H.S. \Seismic response ral networks on super resolved images", Integrated
prediction method for building structures using con- Computer-Aided Engineering, 25(4), pp. 321{333
volutional neural network", Structural Health Moni- (2018).
toring, 27(5), p. e2519 (2020). 101. Reyes, O., Fardoun, H.M., and Ventura, S. \An
89. Zhang, R., Chen, Z., Chen, S., Zheng, J., ensemble-based method for the selection of instances
Buyukozturk, O., and Sun, H. \Deep long short-term in the multi-target regression problem", Integrated
memory networks for nonlinear structural seismic Computer-Aided Engineering, 25(4), pp. 305{320
response prediction", Computers and Structures, 220, (2018).
pp. 55{68 (2020). 102. Hamreras, S., Boucheham, B., Molina-Cabello, M.A.,
90. Wang, J., Zhang, L., Chen, Y., and Yi, Z. \A Bentez-Rochel, R., and Lopez-Rubio, E. \Content-
new delay connection for long short-term memory based image retrieval by ensembles of deep learning
networks", International Journal of Neural Systems, object classi ers", Integrated Computer-Aided Engi-
28(6), 1750061 (2018). neering, 27(3), pp. 317{331 (2020).
91. Gulgec, N.S., Takac, M., and Pakzad, S.N. \Struc- 103. Rokibul Alam, K.M., Siddique, N., and Adeli, H. \A
tural sensing with deep learning: strain estima- dynamic ensemble learning algorithm for neural net-
tion from acceleration data for fatigue assessment", works", Neural Computing and Applications, 32(10),
Computer-Aided Civil and Infrastructure Engineer- pp. 6393{6404 (2020).
ing, 35(12), pp. 1349{1364 (2020).
92. Ahmadlou, M. and Adeli, H. \Enhanced probabilistic Biographies
neural network with local decision circles: a robust
classi er", Integrated Computer-Aided Engineering, Juan Pablo Amezquita-Sanchez graduated from
17(3), pp. 197{210 (2010). University of Guanajuato in 2007 with a BSc degree in
93. Hirschauer, T.J., Adeli, H., and Buford, J.A. Electronic Engineering. He received his MSc degree in
\Computer-aided diagnosis of Parkinson's disease us- Electrical Engineering from University of Guanajuato
ing enhanced probabilistic neural network", Journal and the PhD degree in Mechatronics from the Au-
of Medical Systems, 39(11), p. 179 (2015). tonomous University of Queretaro, Queretaro, Mexico
94. Amezquita-Sanchez, J.P., Adeli, A., and Adeli, H. in 2009 and 2012, respectively. He was a Postdoctoral
\A new methodology for automated diagnosis of Visiting Scholar in The Ohio State University during
mild cognitive impairment (MCI) using magnetoen- 2013-2014. He is currently a full-time Professor at
cephalography (MEG)", Behavioural Brain Research, the Faculty of Engineering, Autonomous University
305, pp. 174{180 (2016). of Queretaro, Campus San Juan del Rio, Queretaro,
95. Amezquita-Sanchez, J.P., Valtierra-Rodriguez, M., Mexico. He is a member of the Mexican National
Adeli, H., and Perez-Ramirez, C. \A novel wavelet Research System (SNI), level 1. He has published
2656 J.P. Amezquita-Sanchez et al./Scientia Iranica, Transactions A: Civil Engineering 27 (2020) 2645{2656

papers in the areas of structural health monitoring, journals Computer-Aided Civil and Infrastructure En-
signal processing, and mechatronics. gineering which he founded in 1986 and Integrated
Computer-Aided Engineering which he founded in
Martin Valtierra-Rodriguez received the BE degree 1993. He has authored over 600 research and scienti c
in Mechatronics Engineering and the ME degree in publications in various elds of computer science, engi-
Electrical Engineering from the University of Guana- neering, applied mathematics, and medicine, including
juato, Guanajuato, Mexico in 2008 and 2010, respec- 16 ground-breaking high-technology books. He is the
tively, and the PhD degree in Mechatronics from the recipient of sixty-seven awards and honors including
Autonomous University of Queretaro, San Juan del four Honorary Doctorates, and Honorary Professorship
Ro, Mexico in 2013. He is currently a Professor at at several Asian and European Universities. In 2005, he
the Faculty of Engineering, Autonomous University was elected Distinguished Member, ASCE: \for wide-
of Queretaro and a member of the Mexican National ranging, exceptional, and pioneering contributions to
Research System (SNI, level 2). His current research computing in civil engineering and extraordinary lead-
interests include signal processing, expert systems, and ership in advancing the use of computing and infor-
applications of mechatronics. mation technologies in many engineering disciplines
throughout the world." He is a member of Academia
Hojjat Adeli received his PhD from Stanford Uni- European, a corresponding member of the Spanish
versity in 1976 at the age of 26. He is currently an Royal Academy of Engineering, a foreign member of
Academy Professor at The Ohio State University where Lithuanian Academy of Sciences and Polish Academy
he held the Abba G. Lichtenstein Professorship for ten of Science, and a Fellow of AAAS, IEEE, AIMBE, and
years. He is the Editor-in-Chief of the international American Neurological Association.