| A | B | C | |
|---|---|---|---|
1 | Filename | Demos featured in | Description |
2 | ml_generative_Gaussian | demo_multiclass_discrim, demo_multiclass_KDE, demo_multiclass_mix, demo_multiclass_NB, demo_multiclass_student | Computes the maximum likelihood Gaussian fit for the data. Targets y are expected to be empty and are ignored. The prediction function returns an [nTest, 1] vector with the likelihood of each data point in the test set. |
3 | ml_generative_KDE | demo_multiclass_discrim, demo_multiclass_KDE | Computes the maximum likelihood kernel density estimate fit for the data. Targets y are expected to be empty and are ignored. The prediction function returns an [nTest, 1] vector with the likelihood of each data point in the test set. |
4 | ml_generative_mixtureGaussian | demo_multiclass_discrim, demo_multiclass_mix | Finds the best fit of k Gaussians to the data using Expectation Maximization. Targets y are expected to be empty and are ignored. The prediction function returns an [nTest, 1] vector with the likelihood of each data point in the test set. |
5 | ml_generative_NB | demo_multiclass_discrim, demo_multiclass_NB | Computes the maximum likelihood Naive Bayes (Gaussian/categorical counts) fit for the data.Targets y are expected to be empty and are ignored. The prediction function returns an [nTest, 1] vector with the likelihood of each data point in the test set. |
6 | ml_generative_student | demo_multiclass_discrim, demo_multiclass_student | Computes the maximum likelihood Student-t fit for the data. Targets y are expected to be empty and are ignored. The prediction function returns an [nTest, 1] vector with the likelihood of each data point in the test set |
7 | ml_multiclass_1v1 | demo_multiclass_v1 | Train C(C-1)/2 binary classifiers for a C-way multiclass problem; each receives the samples of a pair of classes from the original training dataset, and votes on the most probable class |
8 | ml_multiclass_1vA | demo_multiclass_v1 | Train C binary classifiers on C classes |
9 | ml_multiclass_bagging | demo_multiclass_bagging | Classication based on the highest prediction among models trained on bootstrap samples of dataset. |
10 | ml_multiclass_basis | demo_multiclass_basis, demo_multiclass_multiclass_CV | Classification with a basis change. |
11 | ml_multiclass_boosting | demo_multiclass_boosting | Implements a SAMME AdaBoosting algorithm to boost a classifier (see in Zhu, Rosset, Zou, and Hastie 2006) |
12 | ml_multiclass_CV | demo_multiclass_multiclass_CV | This computes the "best" hyper-parameter(s) using cross-validation for classification and regression problems |
13 | ml_multiclass_decisionTree | demo_multiclass_decisions | A decision tree that classifies data into multiple class labels |
14 | ml_multiclass_ECOC | demo_multiclass_ECOC | Relabel classes into binary forms, and use combination of binary classification submodels to predict multiclass labels |
15 | ml_multiclass_GDA | demo_multiclass_discrim | Classification using a Gaussian generative model with shared full covariance matrices to fit each class |
16 | ml_multiclass_kernel_softmax | Fits a classification model by minimizing the kernel softmax loss function | |
17 | ml_multiclass_KNN | demo_multiclass_KNN | An object is classified by a majority vote of its neighbors, with the object being assigned to the class most common among its k nearest neighbors. |
18 | ml_multiclass_logistic | demo_multiclass_bagging, demo_multiclass_basis, demo_multiclass_KNN, demo_multiclass_multiclass_CV | Classification using multinomial logistic regression |
19 | ml_multiclass_MLP | demo_multiclass_boosting | Classification using a multilayer perceptron with softmax loss |
20 | ml_multiclass_stump | demo_multiclass_decisions | Finds the best threshold across all features |
21 | ml_multiclass_SVM | demo_multiclass_SVM | Fits a linear classifier by maximizing the margin using SVM |
22 | ml_multiclass_CNN | demo_multiclass_CNN | Finds optimal set of parameters for simple convolutional neural network in multiclass classification regime. |