Machine Learning–Types of Machine Learning –Machine Learning process- preliminaries, testing Machine Learning algorithms - turning data into Probabilities and Statistics for Machine Learning - Probability theory – Probability Distributions – Decision Theory – Bias, Variance and Tradeoff. Lab Practice: 1. Implement data crawlers Beautiful Soup, Ixml and scrapy 2. Implement data analyzing for a data set using SciPy and generate graph using NetworkX 3. Implement Resource Description Frame work by DBpedia 4. Create one- and two-dimensional random dataset and implement Series and Data Frames in python using slicing methods. 5. Implement the python code to compute the population proportions for the given problem.

Naive Bayes Classifiers - Decision Tree - Training and Visualizing a Decision Tree - Entropy and Understanding cost function and gradient descent - Overfitting and Underfitting - Classification and Regression - Linear Regression - Least Squares – Ridge - Lasso - Polynominal Regression - Logistic Regression: - SVM and Hyperparameter tuning - Implementing SVM using scikit-learn Lab Practice: 1. Implement Linear and Logistics regression using SciKit Learn [May use appropriate library] 2. Implement Stock Price Prediction based on Scikit Learn [May use appropriate library]

Naive Bayes Classifiers - Decision Tree - Training and Visualizing a Decision Tree - Entropy and The CART Training Algorithm - Random Forests, – Ensemble - Boosting - AdaBoost and Gradient Boosting Lab Practice: 1. Predict house prices based on several parameters available in the Housing and Urban Development of any dataset using least squares linear regression. 2. Implement sentiment analysis using random forest optimization algorithm

Clustering- K-means – EM Algorithm- Mixtures of Gaussians – Hierarchical Clustering- K- Nearest Neighbours - Dimensionality Reduction - Linear Discriminant Analysis, Factor Analysis, Principal Components Analysis, Independent Components Analysis. Lab Practice: 1. Build a movie recommendation engine by applying collaborative filtering and topic modelling techniques. The dataset may contain 20 million viewer ratings of 100 movies. 2. Implement Market Segmentation using K-means Clustering. 3. Implement k-Nearest Neighbour algorithm to classify the provided data sets.

Introduction - Single State Case - Elements of Reinforcement Learning – Markov Decision Process - Model Based Learning – Model Free Learning - Temporal Difference Learning – Q Learning Algorithm – Generalization - Partially Observable States. Lab Practice: 1. Implement Q Learning with Linear Function Approximation