-
HW1--Linear Classifiers:
- KD tree (data structure).
- Trainig data set using customised 2D multivariate Gaussian distributions.
- Maximum-likelihood linear least squares classifier.
- Nearest neighbors classification using KD tree.
- Testing above both classifier using overlapped Gaussian distributions.
-
HW2--Bayes Classifiers:
- Naive Bayes (uniform Dirichlet prior) on MNIST data set.
- Naive Bayes (Gaussian kernel) and ROC curve (FPR vd TPR).
- Brute force nearest neighbors search (no KD tree).
-
HW3--Regression on NHSA automobile crash test data:
- Linear regression: polynomial basis functions.
- Linear regression: radial basis functions.
- Bayesian MAP estimates.
- Logistic regression on Iris dataset
-
HW4:
- Gaussian process regression with five-fold cross-validation.
- K-means clustering.
- Hidden Markov Models (forward-backward algorithm).
sayin/Machine_Learning_CS_5783
Python based machine learning codes developed for CS_5783 course
Python