Machine Learning

Implementation of Some of the Machine Learning (ML) Algorithms From Scratch in Python

Homework 1

Problems | Solutions (Report)

• Data Visualization (Source code)
• Univariate Polynomial Regression using Gradient Descent and Normal Equation (Source code)
• Multivariate Polynomial Regression, Stepwise Feature Selection (Backward Elimination), Cook's Distance and DFFITS plots from scratch. Fixing Heteroscedasticity and Multicollinearity. (Source code)

Homework 2

Problems | Solutions (Report)

• Decision Tree (Source code)
• Ensemble Learning - Bagging algorithm using Decision Tree (Random Forest) (Source code)
• KNN using Euclidean and Cosine distance, Confusion Matrix for Multi-class Classification, Optical Recognition of Handwritten Digits (Source code)

Homework 3

Problems | Solutions (Report)

• Implementation of Naive Bayes Algorithm for Spam Email Detection (Source code)

Homework 4

Problems | Solutions (Report)

• Custom kernels for Support Vector Machine (SVM) (Source code)

Homework 5

Problems | Solutions (Report)

• K-means clustering, Outlier detection (Source code)
• DBSCAN, Evaluation Metrics (Accuracy, Entropy, Purity) (Source code)
• Q-learning through Epsilon-Greedy Algorithm (Source code)

Final Project

Problems | Solutions (Report)

• Multi-output Regression (Source code)
• Skin Detection (Source code)
• Credit Approval (Source code)

Author

Rabist - view on LinkedIn

Details

• Course: Machine Learning (CE5501) - MS
• Teacher: Dr. Ehsan Nazerfard
• Univ: Amirkabir University of Technology
• Semester: Fall 2022

License

Licensed under MIT.