Machine Learning Algorithms

Description

In this repository I implement Machine Learning algorithms in Kaggle competitions, preparing the data beforehand through preprocessing, cleaning and feature engineering using Pandas library.

Some algorithms were developed from scratch using Pytorch for gradient descent, and for others I used Scikit-learn models for efficiency. All the algorithms were tested in competitions, obtaining good results.

I've also worked with the Numpy library to prepare the tensors for the Pytorch deep neural networks.

Algorithms

House Prices - Advanced Regression Techniques

• Linear Regression
• Decision Trees
• Random Forests
• Gradient Boosting using XGBoost
• Principal Component Analysis using Linear Regression
• Ridge / Lasso Linear Regression (Both)

Titanic - Machine Learning from Disaster

• Logistic Regression
• K-Nearest Neighbors
• AdaBoost

Digit Recognizer

• Multilayer Perceptron

K-Means Clustering for Heart Disease Analysis

• K-Means Clustering

Prediction of spam with Bayesian model

• Gaussian Naive Bayes