Decompose gbm predictions into feature contributions

Machine Learning approach to predicting comment recommendations based on New York Times articles and comments made on those articles.

Une expérimentation autour d'un petit jeu de données de Churn bancaire

h2o is a ML library that can be used with R and Python. Here are some R examples for supervised and unsupervised methods.

Various variable selection methods are explored

The project involves deciding on the mode of transport that the employees prefer while commuting to office. For this, multiple models such as KNN, Naive Bayes, Logistic Regression have been created and explored to check their model performance metrics. Bagging and Boosting modelling procedures have also been applied to create the models.

A R script that runs Boosted Regression Trees(BRT) on epochs of land use datasets with random points to model land use changes and predict and determine the main drivers of change

Project to produce supervised ML algorithm to predict which customers are likely to leave and produce .Rmd report

R-based project to analyze lyrics entropy by genre and decade. A hand-engineered feature "words-per-unique-word" is introduced and deeply studied. Spotify and Genius APIs are used

The objective of the project is to create a machine learning model. We are doing a supervised learning and our aim is to do predictive analysis to predict median housing price.

This is a Masters project completed by My Team and I using the statistical methodology Markov Chain and Geometric Brownian Motion for trend and closing price prediction

Analysis music on Spotify and predictions of preferences

Modelling Experts Behaviour in Q&A Communities to Predict Worthy Discussion

Building binary predictors on a heavily imbalanced dataset - exercise on policy cross-selling [kaggle]

In this project, exploratory data analysis was used to identify reasons why employees leave and machine learning methods were used predict employee attrition

Detect Credit Card Fraud with Machine Learning in R

This project about the GBM classification model on spam email data set and model optimisation.

Single-cell multi-omic profiling of glioblastoma-associated myeloid cells