ML Projects

Objective

In the following projects, I learned the basics of some machine learning and got started with knowledge in the field. The following projects covered these topics: linear regression, k-means, deep learning: convolutional neural networks, and q-learning. Enjoy!

Projects

Linear Regression Projects

Libraries: SkLearn, Pandas, Seaborn, MatPlotLib, and NumPy

Projects: Two projects related to linear regression were done. One was predicting housing prices in Boston, while the other was predicting the survival of people on the Titanic. Here are some images of the Titanic project below. I used a 70%-30% test and validation sets.

K-Means Model

Libraries: Seaborn, MatPlotLib, SkLearn

Projects: In this project, I had an array of data, and given the array, I created a k-means model to group this data into four separate compartments.

Convolutional Neural Network Project

Libraries: Pandas, Numpy, Matplotlib, Seaborn, Random, and TensorFlow

Projects: Using the fashion MNIST, I made a convolutional neural network that can determine different types of articles of clothing. I used a 30/70 testing and training set. In this project, I used the library Keras from Tensorflow.

Q-Learning

Libraries: Gym, Numpy

Projects: I also did a Q-Learning project here. I used the gym, NumPy, and Bellman's equations to make the computer win a specific taxi game and score the highest possible score.

Google Gym Atari Games (Reinforcement Learning)

Libraries: Keras Sequential Model (Layers: Dense, Dropout, Concetenate), Google Gym, Numpy

Projects: I used reinforcement learning with Atari games. This was acquired from using Google Gym.