MachineLearning

Conventional ML Algorithms

1. Implementation of K-nearest Neighbours (KNN) and Linear Regression

2. Implementation of (1) Mixtures of Gaussians and (2) Logistic Regression for Classification

3. Non-linear regression

   • Regularized Generalized Linear Regression
   • Bayesian Generalized Linear Regression
   • Gaussian Process Regression (Kernel Method)

Modern ML Algorithms (Deep Learning)

4. Convolutional Neural Networks (CNNs) in TensorFlow

   • Experiment with Fully Connected Neural Networks and Convolutional Neural Networks

5. Sequence Models with Recurrent Neural Network (RNN) in PyTorch

   • Encoder Implementation with Linear Hidden Units, GRUs and LSTM Units

     • Application: Predicting which language a name is from based on the spelling

     

   • Decoder Implementation

     • Application: Generating names from languages

     

   • Natural Language as Sequences: Sequence to Sequence Model (with Attention) Implementation

     • Application: Translating from one language to another