https://github.com/jasontanx/machine-learning-employee-attrition
An employee attrition prediction (machine learning) project
https://github.com/jasontanx/machine-learning-employee-attrition
data-science employee-attrition machine-learning rprogramming
Last synced: 3 months ago
JSON representation
An employee attrition prediction (machine learning) project
- Host: GitHub
- URL: https://github.com/jasontanx/machine-learning-employee-attrition
- Owner: jasontanx
- Created: 2023-02-08T14:56:30.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-02-12T11:17:38.000Z (over 2 years ago)
- Last Synced: 2025-02-01T02:19:53.833Z (5 months ago)
- Topics: data-science, employee-attrition, machine-learning, rprogramming
- Language: R
- Homepage:
- Size: 471 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# machine-learning-employee-attrition 🧑💼
**Topic:**
Building a Reliable Predictive Model to Predict Employee Attrition
*R-Programming will be the coding language used in this project*
No | Dataset | Information
--- | --- | ---
1 | URL | https://www.kaggle.com/rushikeshghate/capstone-projectibm-employee-attrition-prediction?select=IBM+HR+Data+new.csv
2 | Dataset Name | IBM Employee Attrition
3 | File Type | csv file
4 | Observation | 23,436
5 | Features | 37
6 | Data label |Staying employee as “Current Employee” while resigned employee as “Voluntary Resignation”.
# Introduction
## Why is employee attrition an issue?
- High turnover rates / employee attrition are becoming a huge issue for organisations
- The Great Attrition --> [Click here for more information about The Great Attrition](https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-great-attrition-is-making-hiring-harder-are-you-searching-the-right-talent-pools)
- High cost of hiring + training + time-taken for recruitment & onboarding
## What is the problem statement of the project?
- Challenges for organisations to retain their current workforce
- Lack of reliable machine learning tools to predict employee attrition
## Aims & Objectives (What do I aim to achive?) 🌟
**The Aims**
- The overall aim of this project is to build a predictive machine learning model that would play a vital part in HRM by accurately predicting employee attrition
**The Objectives**
- To obtain employee attrition dataset with more observations as compared with previous work
- To identify the key features that could affect prediction of employee attrition
- To build a model (and evaluate model performance) that could accurately predict employee attrition
# Data Preparation
- Dataset exploration 🔍
- Data selection & cleaning 🧹
- Removing irrelavant features
- Convert data structure to the suitable ones (*cases where categorical variables are loaded as character instead of factor in RStudio*)
- Missing values & imputation (*Let's replace the missing values!*)
- Feature selection (*Does some variables really significantly contribute to the prediction of employee attrition?*)
- Class distribution of the target variable (*It seems there might be a huge class imbalance issue*) --> Data Re-sampling needed!
- Click here to know more about oversampling --> [What is oversampling or undersampling in machine learning?](https://machinelearningmastery.com/random-oversampling-and-undersampling-for-imbalanced-classification/)
- Exploratory Data Analysis (EDA) 🧮
- Univariate analysis
- Bivariate analysis
- One-Hot Encoding / Label Encoding 🖥️
# Model Implementation
- Logistic Regression (LR)
- Decision Tree (DT)
- Naive Bayes (NB)
# Performance Evaluation
- DT performed the best as compared with other models experimented
- Model perofmrance for DT is consistent
- The dataset used was the oversampling dataset without the issue of class imbalance
# Recommendations for Future Work
- Dataset with more observations
- Build predictive model with real HR dataset (*Current dataset is a fictional dataset provided by IBM*)
- Include more variables (*Such as work from home, hybrid or full physical*)
- Different feature selection algorithm (*Boruta algorithm was used in this project*)
# Further Information About the Topic
**Employee Attrition Prediction with Python**
- [Employee Attrition Rate Prediction Using Machine Learning](https://www.enjoyalgorithms.com/blog/attrition-rate-prediction-using-ml)
- [Employee Attrition Prediction - A Guide](https://www.analyticsvidhya.com/blog/2021/11/employee-attrition-prediction-a-comprehensive-guide/)