Attrition Prediction of IBM employees

Description

This project introduces a module designed to forecast the attrition rate, with a specific focus on predicting voluntary employee departures from IBM. In the dynamic landscape of service-oriented companies like IBM, where the expertise of skilled professionals is an invaluable asset, the sudden departure of such personnel can pose a substantial risk to organizational stability. Harnessing the capabilities of Machine Learning and implemented with Python, our project employs the Logistic Regression algorithm for attrition prediction. Central to our approach is the pivotal role played by data analysis, guiding us through the nuanced exploration of this critical organizational challenge.

Languages and Utilities Used

• Python
• Google Colab

Output

In this project, our aim was to develop a predictive model for employee attrition, a critical concern for organizations like IBM. Leveraging machine learning techniques, particularly logistic regression, and employing Python as the programming language, we embarked on an insightful journey to understand and predict the attrition rate. Google Colab was main environment we used for this project.The logistic regression model exhibited an accuracy of approximately 85%, with a precision of 85% for non-attrition cases. However, it struggled to effectively predict attrition, as reflected in the low recall and F1-score for the attrition class. This project lays the groundwork for predicting employee attrition, providing valuable insights for organizational management. Continued refinement and exploration are encouraged to evolve the model into a robust tool for proactive HR management.

Program walk-through:

Project Architecture

Results: F1 score