AI Learning Hub for Machine Learning, Deep Learning, Computer Vision and Statistics

High Dimensional Discriminant Analysis in R ✨

Regularized discriminant analysis in Julia.

This module allows users to analyze k-means & hierarchical clustering, and visualize results of Principal Component, Correspondence Analysis, Discriminant analysis, Decision tree, Multidimensional scaling, Multiple Factor Analysis, Machine learning, and Prophet analysis.

Function preserving projection (FPP), a linear projection technique for capturing interpretable patterns of high-dimensional functions

This is a scanner designed to recognise DNA motifs within a long stretch of DNA. It uses two models for discrimination, one model representing the target and the second model representing the background.

A comprehensive MATLAB-based machine learning project covering data preprocessing, dimensionality reduction, supervised and unsupervised learning algorithms, model evaluation, and validation techniques with practical examples.

Multi-distributional Discriminant Analysis using Generalised Linear Latent Variable Modelling in R ⭐

Проведение бинарной и многоклассовой классификаций эмоций людей на фотографиях

The code for Roweis Discriminant Analysis (RDA) and Kernel RDA methods

R package DiscriMiner

Breast cancer classification and evaluation of classifiers

Code for the paper E. Raninen and E. Ollila, “Coupled regularized sample covariance matrix estimator for multiple classes,” in IEEE Transactions on Signal Processing, vol. 69, pp. 5681–5692, 2021, doi: 10.1109/TSP.2021.3118546.

knn, Regression (LASSO, Ridge), Logistic, Principal component Analysis (PCA), Discriminant Analysis (LDA, QDA), Trees, Random Forest, Boosting

Code for performing LDA-style procedure for when data distributions follow the exponential family distribution.

Materiales de las clases prácticas de AID y Aprendizaje Automático

The software package SiteGA for de novo motif search in ChIP-seq data

FEMDA: Robust classification with Flexible Discriminant Analysis in heterogeneous data. Flexible EM-Inspired Discriminant Analysis is a robust supervised classification algorithm that performs well in noisy and contaminated datasets.

Performed statistical-EDA and normalization analysis on digitized mass images with 10 nuclei features (radius, texture) Predicted malignant - benign cancer using Logistic, LDA-QDA, KNN, Lasso-Ridge classifiers with 0.89, 0.88, 0.92, 0.96 and 0.97 accuracies respectively along with decision boundaries and ROC curves

Code for ICASSP paper. FEMDA: Robust classification with Flexible Discriminant Analysis in heterogeneous data. Flexible EM-Inspired Discriminant Analysis is a robust supervised classification algorithm