The Repository containing notebooks of nearly all major ML algorithms implemented from scratch. This repo was a part of 3116 - Programming Machine Learning being taught at the University of Hildesheim
This contains SLE algos (Gaussian elimination), optimization algos (GD, Newton, SGD, CD), step-length algos (backtracking, bold-driver, RMSProp), forward/backward search with AIC/BIC, grid/randomized search with K-Fold, regularization (L1, L2, Elastic Net, Dropout), losses/fit (MSE, cross-entropy, case weights, MCR, Information gain), Linear/logistic regression, LDA/QDA, KNN (vanilla, LSH), decision trees (with gradient boosted), MLP, SVM, Naïve Bayes, K-Means, Gaussian Mixtures, NLP (TF-IDF, BOW, CBOW model, N-Gram language model)