This repository contains projects assigned by Udacity in their Machine Learning Engineer Nanodegree course. There are 5 projects in total.
This project explores data to estimate the survival rate of passengers on the Titanic ship. This project uses bare data manipulation without any high-level algorithm.
Framework Credits: Udacity
This project focuses on estimating housing prices in Boston using Decision Tree Regression. An optimal model is also found by performing a Grid-search cross-validation over certain parameters of the regression model.
Framework Credits: Udacity
This project demonstrates understanding of Supervised Classification. Support Vector Machines (SVM), Decision Tree Classifier and Gaussian Naive Bayes Classifiers have been explored. A grid-search cross-validation is performed for the SVM and the optimal model's performance was noted.
Framework Credits: Udacity
This project focuses on Unsupervised Learning algorithms such as Dimensionality Reduction using Principal Component Analysis (PCA), Outlier Detection and Clustering. The goal of the project is to uncover hidden segmentation of types of customers based on purchasing patterns.
Framework Credits: Udacity
This project demonstrates the application of Reinforcement Learning - specifically Q-Learning to a simulated smartcab in a simplified world (grid). Over time, the agent (smartcab) learns to obey traffic rules, learns to avoid collisions with other cars and simultaneously learns to navigate to the goal position in the shortest possible path.
Framework Credits: Udacity