employee-salary-prediction

KO | EN

gitlite — search

#swift #list #large-language-models #rag #gemini #http #command-line #linux #javascript #ai

employee-salary-prediction

(★ 9)

Machine learning system for predicting employee salaries with 86.7% accuracy. Includes data preprocessing, model training, evaluation, and prediction interfaces.

File Explorer

ZIP Download

- config.py
- - .gitkeep
  - processed_data.csv
- - .gitkeep
  - demo_employee_data.csv
  - employee_data.csv
- .gitkeep
- best_model.joblib
- best_salary_model.joblib
- decision_tree_model.joblib
- linear_regression.joblib
- preprocessor.joblib
- random_forest.joblib
- random_forest_model.joblib
- ridge_regression.joblib
- .gitkeep
- evaluation_report.txt
- model_results.csv
- prediction_demo_results.csv
- quick_evaluation_results.csv
- scenario_comparison.csv
- simple_predictions.csv
- simple_scenarios.csv
- __init__.py
- data_analysis.py
- data_generator.py
- data_preprocessing.py
- model_evaluation.py
- model_training.py
- prediction_interface.py
- quick_evaluation.py
- simple_predictor.py
- simple_training.py
- test_prediction.py
.gitignore
API_REFERENCE.md
demo.py
LICENSE
quick_demo.py
README.md
requirements.txt
USAGE_GUIDE.md

# Use via CDN

jsDelivr

jsDelivr serves any public GitHub repository as a CDN with zero setup. Pick a version and a file to get a ready-to-paste link and snippet.

Version

File

Link

Example

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

A comprehensive machine learning system for predicting employee salaries based on experience, education, role, industry, and location data.

## 🎯 Problem Statement

HR departments need data-driven salary recommendations to maintain competitive and fair compensation structures. This system provides accurate salary predictions using machine learning to help with:

- Setting competitive salary offers
- Evaluating compensation equity
- Planning salary budgets
- Career progression planning

## 🚀 Solution Overview

The system uses regression models to predict salaries based on:

- **Demographics**: Age, years of experience
- **Education**: Education level, certifications count
- **Job Information**: Job title, industry, location, employment type
- **Performance**: Performance rating

### Key Features

- **Multiple ML Models**: Linear Regression, Ridge Regression, Random Forest
- **Feature Engineering**: Creates 45+ features from 9 input features
- **Model Evaluation**: Comprehensive performance analysis
- **Prediction Interface**: Individual and batch salary predictions
- **Scenario Analysis**: Compare salary impact of different career changes
- **Visualizations**: Charts and plots for insights

## 📊 Performance Metrics

Our best model (Ridge Regression) achieves:

- **RMSE**: $19,924 (prediction error)
- **R² Score**: 0.8669 (explains 86.7% of salary variance)
- **MAE**: $15,548 (mean absolute error)
- **MAPE**: 21.0% (mean percentage error)

## 🛠️ Tech Stack

- **Python 3.8+**
- **scikit-learn** - Machine learning models
- **pandas** - Data manipulation
- **numpy** - Numerical computing
- **matplotlib/seaborn** - Data visualization
- **joblib** - Model serialization

## 📁 Project Structure

```
Employee Salary Prediction/
├── config/
│   ├── config.py              # Project configuration
│   └── __pycache__/
├── data/
│   ├── raw/
│   │   └── employee_data.csv  # Generated dataset
│   └── processed/
│       └── processed_data.csv # Preprocessed data
├── models/
│   ├── best_salary_model.joblib    # Best model package
│   ├── linear_regression.joblib    # Linear regression model
│   ├── ridge_regression.joblib     # Ridge regression model
│   ├── random_forest.joblib        # Random forest model
│   └── preprocessor.joblib         # Data preprocessor
├── outputs/
│   ├── plots/                      # Generated visualizations
│   ├── model_results.csv           # Model comparison results
│   └── evaluation_report.txt       # Detailed evaluation report
├── src/
│   ├── data_generator.py           # Synthetic data generation
│   ├── data_preprocessing.py       # Data preprocessing pipeline
│   ├── model_training.py           # Model training system
│   ├── model_evaluation.py         # Model evaluation system
│   ├── prediction_interface.py     # Advanced prediction interface
│   ├── simple_predictor.py         # Simple prediction interface
│   └── simple_training.py          # Simplified training script
├── notebooks/                      # Jupyter notebooks (optional)
├── requirements.txt                # Python dependencies
└── README.md                       # This file
```

## 🚀 Quick Start

### 1. Installation

```bash
# Clone the repository
git clone <repository-url>
cd employee-salary-prediction

# Install dependencies
pip install -r requirements.txt
```

### 2. Generate Data and Train Models

```bash
# Generate synthetic dataset and train models
python src/simple_training.py
```

This will:
- Generate 10,000 employee records
- Train multiple regression models
- Evaluate model performance
- Save trained models

### 3. Make Predictions

```bash
# Run prediction demo
python src/simple_predictor.py
```

This will demonstrate:
- Individual salary predictions
- Batch predictions for multiple employees
- Scenario analysis (career progression impacts)

## 💡 Usage Examples

### Basic Salary Prediction

```python
from src.simple_predictor import SimpleSalaryPredictor

# Initialize predictor
predictor = SimpleSalaryPredictor()
predictor.load_model()

# Define employee profile
employee = {
    'age': 28,
    'education_level': 'Bachelor',
    'years_experience': 5,
    'job_title': 'Software Engineer',
    'industry': 'Technology',
    'location': 'San Francisco',
    'employment_type': 'Full-time',
    'performance_rating': 4.2,
    'certifications_count': 2
}

# Get salary prediction
result = predictor.predict_salary(employee)
print(f"Predicted Salary: ${result['predicted_salary']:,.0f}")
# Output: Predicted Salary: $139,937
```

### Batch Predictions

```python
employees = [
    {
        'age': 28, 'education_level': 'Bachelor', 'years_experience': 5,
        'job_title': 'Software Engineer', 'industry': 'Technology',
        'location': 'San Francisco', 'employment_type': 'Full-time',
        'performance_rating': 4.2, 'certifications_count': 2
    },
    {
        'age': 35, 'education_level': 'Master', 'years_experience': 12,
        'job_title': 'Senior Data Scientist', 'industry': 'Finance',
        'location': 'New York', 'employment_type': 'Full-time',
        'performance_rating': 4.5, 'certifications_count': 4
    }
]

results = predictor.batch_predict(employees)
print(results[['name', 'predicted_salary']])
```

### Scenario Analysis

```python
base_employee = {
    'age': 28, 'education_level': 'Bachelor', 'years_experience': 5,
    'job_title': 'Software Engineer', 'industry': 'Technology',
    'location': 'San Francisco', 'employment_type': 'Full-time',
    'performance_rating': 4.2, 'certifications_count': 2
}

scenarios = [
    {
        'description': 'Get Master\'s Degree',
        'changes': {'education_level': 'Master'}
    },
    {
        'description': 'Move to NYC',
        'changes': {'location': 'New York'}
    },
    {
        'description': 'Promotion to Senior',
        'changes': {'job_title': 'Senior Developer', 'years_experience': 8}
    }
]

comparison = predictor.compare_scenarios(base_employee, scenarios)
print(comparison[['scenario', 'predicted_salary', 'percent_change']])
```

## 📈 Model Performance

### Model Comparison

| Model | RMSE | MAE | R² Score | MAPE (%) |
|-------|------|-----|----------|----------|
| Ridge Regression | $19,924 | $15,548 | 0.8669 | 21.03% |
| Linear Regression | $19,929 | $15,551 | 0.8668 | 21.03% |
| Random Forest | $20,877 | $15,560 | 0.8539 | 18.37% |

### Key Insights

- **Ridge Regression** performs best overall with lowest RMSE
- All models achieve **R² > 0.85**, indicating excellent fit
- **Random Forest** has lower MAPE but higher RMSE
- Models are robust across different salary ranges

### Feature Importance (Random Forest)

Top 10 most important features:
1. Years of experience
2. Education level
3. Job title (Senior roles)
4. Location (San Francisco, New York)
5. Industry (Technology, Finance)
6. Performance rating
7. Age
8. Certifications count
9. Experience per age ratio
10. High-tech industry flag

## 🎛️ Advanced Features

### Data Preprocessing

The system includes sophisticated preprocessing:

- **Feature Engineering**: Creates 45 features from 9 inputs
- **Categorical Encoding**: One-hot encoding for nominal, ordinal for ordered
- **Scaling**: StandardScaler for numerical features
- **Outlier Handling**: IQR-based outlier capping
- **Missing Value Imputation**: Median/mode imputation

### Model Evaluation

Comprehensive evaluation system:

- **Cross-Validation**: 5-fold CV with multiple metrics
- **Residual Analysis**: Normality tests and distribution analysis
- **Prediction Intervals**: Confidence ranges for predictions
- **Feature Importance**: Model interpretability analysis

### Visualization

Rich visualization capabilities:

- Model performance comparisons
- Residual plots and distribution analysis
- Feature importance charts
- Salary prediction scenarios
- Industry and location comparisons

## 📊 Sample Predictions

### Real-World Examples

| Profile | Location | Education | Experience | Predicted Salary |
|---------|----------|-----------|------------|-----------------|
| Software Engineer | San Francisco | Bachelor | 5 years | $139,937 |
| Senior Data Scientist | New York | Master | 12 years | $213,251 |
| Junior Developer | Austin | Bachelor | 1 year | $83,142 |

### Scenario Impact Analysis

For a Software Engineer in San Francisco:

| Scenario | Salary Change | % Impact |
|----------|---------------|----------|
| Get Master's Degree | +$24,371 | +17.4% |
| Move to NYC | -$9,766 | -7.0% |
| Promotion to Senior | +$31,598 | +22.6% |
| All Improvements | +$54,565 | +39.0% |

## 🔧 API Reference

### SimpleSalaryPredictor Class

```python
class SimpleSalaryPredictor:
    def load_model() -> bool
    def predict_salary(employee_data: Dict) -> Dict
    def batch_predict(employees_list: List[Dict]) -> pd.DataFrame
    def compare_scenarios(base_employee: Dict, scenarios: List[Dict]) -> pd.DataFrame
```

### Required Input Fields

```python
employee_data = {
    'age': int,                    # Employee age (22-65)
    'education_level': str,        # 'High School', 'Associate', 'Bachelor', 'Master', 'PhD'
    'years_experience': float,     # Years of work experience
    'job_title': str,              # Job title (see supported titles below)
    'industry': str,               # Industry (see supported industries below)
    'location': str,               # Work location (see supported locations below)
    'employment_type': str,        # 'Full-time', 'Part-time', 'Contract'
    'performance_rating': float,   # Performance rating (1.0-5.0)
    'certifications_count': int    # Number of professional certifications
}
```

### Supported Values

**Job Titles:**
- Junior Developer, Software Engineer, Senior Developer
- Data Scientist, Senior Data Scientist, ML Engineer
- Team Lead, Manager, Director, VP
- Product Manager, DevOps Engineer, Designer, QA Engineer

**Industries:**
- Technology, Finance, Healthcare, Education
- Manufacturing, Government, Consulting, Retail

**Locations:**
- San Francisco, New York, Seattle, Boston
- Austin, Chicago, Denver, Atlanta, Phoenix, Remote

## 🧪 Testing

Run the test suite:

```bash
# Test prediction functionality
python src/test_prediction.py

# Run evaluation system
python src/quick_evaluation.py

# Test full pipeline
python src/simple_training.py
```

## 📝 Configuration

Edit `config/config.py` to modify:

- Dataset size and random seed
- Model parameters and CV folds
- Feature lists and salary ranges
- Output paths and directories

## 🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests for new functionality
5. Submit a pull request

## 📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

## 🙏 Acknowledgments

- Built with scikit-learn and pandas
- Inspired by real-world HR challenges
- Uses synthetic data for privacy protection

## 📞 Support

For questions or issues:

1. Check the documentation above
2. Review the example scripts in `src/`
3. Run the test scripts to verify installation
4. Create an issue with detailed information

## 🔮 Future Enhancements

Potential improvements:

- **Deep Learning Models**: Neural networks for complex patterns
- **Real-Time Prediction API**: REST API for integration
- **Geographic Expansion**: More locations and regions
- **Industry-Specific Models**: Specialized models per industry
- **Confidence Intervals**: Statistical prediction intervals
- **Model Monitoring**: Performance tracking over time

---

**Happy Predicting!** 🎯💰

employee-salary-prediction

Owner Repos

File Explorer

# Use via CDN

Was this content helpful?

Recent Feedback

employee-salary-prediction

Owner Repos

File Explorer

# Use via CDN

Was this content helpful?

Recent Feedback

Download README