A fast xgboost feature selection algorithm

All Relevant Feature Selection

Migrated to GitLab

Feature selection on a Bitcoin Dataset

A Human Resource Analytics project that addresses the question "Why do employees leave?". Also helps us better understand factors for employee satisfaction

The feature of interest is whether or not a customer buys a caravan insurance, based on socio-demographic factors and ownership of other insurance policies; and to build profile of a typical customer.

Analytical tool to help the company decide whether the employee will stay or not

This web application is motivated by Baymax of the animated movie Big Hero 6. It detects Valvular Heart disorder i.e. damage or defect in one of the four heart valves. On the Machine Learning side, I have used AutoML from the deep learning platform H2O. And the interactive application is build in RShiny.

Predictions of clinical variables with sensible feature selection. Includes: Elastic net, BORUTA, hierarchical regression

Modelling and prediction of default + deployment via AWS Sagemaker

All-relevant feature selection with Boruta, Shiny and R

Classification model using R to predict the biodegradability of various chemicals

ML-based prediction of NSCLC recurrence with gene expression data

Project for IST687 Applied Data Science- School of Information Studies, Syracuse University

Identify the most important variables that contribute to the level of sustainability of schools using supervised machine learning. Use the importance to assign a weight to each characteristic and calculate a sustainability index for each school.

This repository proposes several methods to infer relevant associations within a set of features.

"Heart Attack Risk Prediction" uses machine learning to estimate the likelihood of a heart attack based on user-provided data like physical attributes, symptoms, and medical history. This system enables remote screening, identifying high-risk individuals, and easing medical system burdens by providing early, data-driven health risk assessments.

This repository contains an ML workflow to predict house prices in Ames, Iowa. This project work is carried out under the Machine Learning module of the GeoDSc track of the Copernicus Master in Digital Earth.

Sales Prediction for Rossmann Pharmaceutics database using Machine Learning regression modeling

Machine learning model developed in R to predict which water pumps across Tanzania are not working properly