ML projects, application and implementation
these projects were done in order to implemenet what we learn and includes kaggle competetions also.
ML assignment-1 contains basics of ML i.e. linear and logistic regresssions ( implementation from scratch) along with their documentations as well as some fascinating results we got while performing some exercises through out the course. It includes evaluation of Bias and variance, views upon the learning rate and many small techniques used to avoid overfitting and underfitting of models.
ML assignment-2 contains ML algorithms like descision trees, Naive bayes and gaussian naive bayes (implemented from scratch) as well as comparision of them with the SKlearn libraries. Also, it includes understanding about concepts of TSNE, SNE, SVD, PCA and stratified sampling as steps for dimensionality reduction.
ML assignment-3 contains scratch implementation of NN model with sigmoid, softmax, relu, linear and tanh as activation functions and much more which can be find in the ductmentation of ML assignment-3, alsp we use dpytorch implementation of CNN along with pre-trained alex net model to classify images based on the data. Alexnet model output layer which contain 1000 nodes is used as input layer to NN with 256 128 64 hidden layers with output of 2 ( +1/-1).