hr-analytics-attrition-prediction

📌 Project Overview

Employee attrition is a major challenge for organizations as it increases hiring costs and affects productivity. This project analyzes historical HR data to understand the main causes of employee resignation and builds a machine learning model to predict future attrition.

The project combines Python (EDA + ML) and Power BI (Interactive Dashboard) to deliver actionable HR insights.

🎯 Objective

• Analyze key factors influencing employee attrition
• Build a classification model to predict attrition
• Explain model predictions using SHAP
• Develop an interactive HR dashboard for decision-making

🛠 Tools & Technologies Used

• Python
• Pandas & NumPy
• Seaborn & Matplotlib
• Scikit-learn (Logistic Regression, Decision Tree)
• SHAP (Model Explainability)
• Power BI (Dashboard & DAX)

📊 Exploratory Data Analysis (EDA)

Performed analysis on:

• Department-wise attrition
• Salary bands
• Overtime impact
• Job role analysis
• Years at company vs attrition
• Promotion & job satisfaction trends

Key insights identified major risk factors contributing to employee resignation.

🤖 Machine Learning Model

Built classification models:

• Logistic Regression
• Decision Tree Classifier

Evaluation Metrics:

• Accuracy Score
• Confusion Matrix

The model predicts whether an employee is likely to leave the organization.

🔍 SHAP Analysis

Applied SHAP (SHapley Additive Explanations) to:

• Identify most influential features
• Improve model interpretability
• Understand feature impact on attrition probability

Major factors identified:

• Overtime
• Monthly Income
• Years at Company
• Job Role

📈 Power BI Dashboard

Developed an interactive dashboard including:

• KPI Cards (Total Employees, Attrition Rate)
• Donut Charts (Attrition Distribution)
• Department-wise Analysis
• Job Role Analysis
• Years at Company Trend
• Decomposition Tree
• Key Influencers Visual

The dashboard provides executive-level HR insights.

📌 Project Deliverables

• Trained ML Model
• Model Accuracy Report & Confusion Matrix
• SHAP Summary Analysis
• Power BI Dashboard
• Project Report (PDF)

🚀 Key Outcomes

• Identified key drivers of employee attrition
• Built predictive model with strong performance
• Created interactive visualization for HR decision-making
• Demonstrated application of ML + Business Intelligence integration tructure (Example)