Machine-Learning-Course-Assignments

Assignments of Machine Learning Graduate Course under supervision of Dr. Ehsan Nazerfard - Spring 2021

HW1 :

problem1:

• implementation of linear Regression with gradient descent algorithm
• fitting polynomial line with different degrees on signal data (univariate)

plots:

Fitting Plots:

Pace Size vs Iterations plot :

Error plot of train and test for degree 7:

Problem2:

• getting familiar with concepts such as: correlation, multivariate linear Regression, feature selection, Regularization, Normal Equation, ...

plot of Data:

Correlation Diagram:

Error vs Iterations:

Pace Size vs Iterations:

HW2 :

problem1:

• implementaion of KNN on MNIST dataset and reporting accuracy, confusion matrix and plotting pictures.

part of Results:

problem2:

• implementation of Naive Bayes Algorithm on Wine dataset, reporting accuracy using 6-fold-cross-validation and plotting ROC curve.

ROC curve:

problem3:

• implementaion of Logistic Regression (one-vs-all) on MNIST dataset and plotting results.

part of Results:

HW3 :

problem1:

• SVM using python libraries (linear, polynomial, RBF and sigmoid kernels)
• reporting accuracy and F1-measure

problem2:

• implementation of Ensemble methods: RandomForest, AdaBoost, Gradient Boosting
• reporting accuracy and confusion Matrix.

HW4 :

problem1:

• implementation of KMeans clustering Algorithm

Results on first dataset:

Results on second dataset:

Results on third dataset:

problem2:

• implementation of DBScan clustering Algorithm
• reporting purity measure

Results on first dataset:

Results on second dataset:

Results on third dataset:

problem3:

• implementation of Hierarchical clustering Algorithm

Results:

problem4:

• implementation of Reinforcement Learning (Deep Q Network) in Mountain Car Environment.

Plots: