Employee Attrition Using Gradient Boosting

Employee turnover and attrition has become a global concern for every company and organization that seize to exist. This research intends to analyze employees' data and accurately predict the significant factors responsible for attrition. To be precise it is a machine learning prediction model that uses a gradient boosting algorithm to pre-process, analyze and predict the factors responsible for attrition. Gradient boosting is machine learning technique that combines multiple predictive models like decision trees to form a more accurate model by sequentially training each model such that it corrects the error of its predecessors. CatBoost is an open-source machine learning library that uses gradient boosting to work with categorical data. It uses decision trees for classification and regression. It is designed to handle complex problems without much data preprocessing. It will be shown in this research how this project acquired the datasets, pre-processed it and how machine learning used data to work with the gradient boosting algorithm and predict the important factors.