Scribd
Upload
35 views

Regression - Classification:: When Is Categorical

Machine learning algorithms can be categorized as regression, classification, clustering, dimensional reduction, or neural networks. Regression algorithms such as linear regression and logistic regression model relationships between independent and dependent variables. Classification algorithms such as support vector machines find boundaries between classes of data. Clustering algorithms such as k-means group similar data points together. Dimensional reduction techniques like principal component analysis reduce the dimensionality of data. Neural networks like convolutional neural networks can learn complex patterns from data.

Uploaded by

Jordan Chizick
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
35 views

Regression - Classification:: When Is Categorical

Machine learning algorithms can be categorized as regression, classification, clustering, dimensional reduction, or neural networks. Regression algorithms such as linear regression and logistic regression model relationships between independent and dependent variables. Classification algorithms such as support vector machines find boundaries between classes of data. Clustering algorithms such as k-means group similar data points together. Dimensional reduction techniques like principal component analysis reduce the dimensionality of data. Neural networks like convolutional neural networks can learn complex patterns from data.

Uploaded by

Jordan Chizick
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Function: Model the relation between the independent

variables and the dependent variable

Function: Automatically group data into clusters, to find


Linear Regression : use when dependent variable is structure inside the data
real value
Clustering
Regression - Classification Kmean: Group all data points into clusters with closest data
Logistic Regression : when dependent variable is to cluster's center
categorical.
Spectral Clustering: User eigenvalues of similarity matrix to
cluster data in lower dimension.
Classification: : find a hyperplane that has largest margin to the
different classes

Regression: : find a hyperplane that is close to all the data points Support Vector Machine (SVM)
in a range Epsilon
Function: reduce dimension of data, easier to visualize,
extract features best represent the data

Convolutional Neural Network :User convolution layer to Principal Component Analysis (PCA): Unsupervised
find features that reside in local group of data. Usually
applied to image processing, currently the best.
Dimensional Reduction Linear
Machine Learning
Linear Discriminant Analysis (LDA): Supervised
Recurrent Neural Network (RNN) : Connections of the
nodes in model form a directed cycle. Can be used to model
dynamic memory when learning.

Neural Network
Auto Encoder : Input and output layers has the same
number of nodes, hidden layer has fewer nodes to compress Locally-linear Embedding (LLE): Try to preserve distances to
data with fewer info only other local objects.

Non-linear - Manifold learning


Function: user layers of many computational nodes, model Multidimensional Scaling (MDS): Try to preserve between-
the neuron in human brain, to learn features from data and object distances of all objects.
calculate the output.
It's very powerful because many layers can represents very
complex model. Function: Use Statistical method to update the probability of
hypothesis

Bayesian Inference : User Bayes theorem, use prior


probability to update posterior probability

Hidden Markov Model (HMM) : can be presented as


simplest dynamic Bayesian network, model is Markov
process with unobserved state*.

Statistical Model Conditional Random Field (CRF): used for structure


prediction, can take context into account, popular in NLP.

Markov Random Field (MRF) : a random field which


satisfies 3 Markov properties

You might also like