Machine-Learning-and-Pattern-Recognition-MO444

Repository for the course "Machine Learning and Pattern Recognition" - MO444 at University of Campinas

Final Assignment:

Reinforcement Learning (Policy search), Deep Learning (CNN and Transfer Learning), Image identification (YOLO) and Stochastic Optimization (Genetic algorithm) techniques were used to optimize the policy of the OpenGymAI Lunar-Lander-V2 Environment

Development made with help of Tensorflow and DEAP packages.

video presentation: https://www.youtube.com/watch?v=9cL4qHz2VlQ

Assignment 1: Linear Regression

Implementation of Linear regression and Bayesian regression using LSE and MLE. Comparison between the optima closed formula and gradien descent.

Assingment 2: Classification

Predict the model of a camera based on the content of their photographs. In other words, given an image, from which camera model did it come from?

Here, logistic regression and many image processing techniques have been applied.

Assignment 3: Clustering and NLP

Implementation of the clustering technique K-Means and different Natual Language Processing (NLP) methods in order to cluster and classify articles headlines according to their subject

Assignment 4: Deep Learning

In this assignment, different techniques involving Deep Learning have been implemented and tested in order to classify images of dogs according to their breeds.