💼 Salary Predictor – Machine Learning Web App

A machine learning-based web application that predicts employee salaries in USD based on job-related parameters using Random Forest Regression. Built with Streamlit and trained on a dataset of over 21,000 IT job roles.

🔗 Live App: https://paypredictor.streamlit.app/
📦 GitHub Repo: https://github.com/PrathmeshSose/salary-predictor Report : Salary_Prediction_Project.pdf

---

📊 Project Overview

This project uses a machine learning model to predict salaries based on:

• Work year
• Experience level
• Employment type
• Job title
• Employee residence
• Remote work ratio
• Company location
• Company size

It uses a dataset sourced from a public GitHub repository and scholarly research, with no missing values and more than 21,000 job records.

---

📁 Dataset Information

• Dataset Name: IT Job Salary Overview
• Records: 21,333
• Features: 11
• Source: damalialutfiani/Salary-Dataset-Project
• Top Job Title: Data Scientist (4,996 records)
• Most Common Currency: USD (19,912 records)
• Avg Salary (USD): $151,037
• Range: $15,000 to $800,000
• Missing Values: None

---

⚙️ Machine Learning Model

✅ Algorithms Explored

• Linear Regression
• Decision Tree Regressor
• Random Forest Regressor ✅
• Gradient Boosting (XGBoost/LightGBM)
• Support Vector Regression (SVR)
• Neural Networks (for large datasets)

📦 Libraries Used

pandas
numpy
scikit-learn
matplotlib
seaborn
streamlit
joblib

---

🧠 Model Training Steps

1. Import necessary libraries
2. Load and explore dataset (final_salaries.csv)
3. Drop unnecessary columns (salary, currency)
4. Encode categorical features with OrdinalEncoder
5. Split data into training and test sets
6. Train RandomForestRegressor model
7. Evaluate using:
   • Mean Absolute Error (MAE)
   • R² Score
8. Save trained model using joblib

---

🖥️ Streamlit App Features

💡 Sections:

• Dataset – View raw job data and shapes
• Graphs – EDA: experience level, salary vs. remote, company size
• Prediction – Input parameters to get live salary prediction 💵

🌐 Deployment:

• Hosted on Streamlit Cloud
• GitHub repo connected to auto-deploy on commit

---

🔧 How to Run Locally

# 1. Clone this repo
git clone https://github.com/PrathmeshSose/salary-predictor
cd salary-predictor

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

# 3. Run the Streamlit app
streamlit run app.py

---

📂 Project Structure

salary-predictor/
├── app.py                # Streamlit application
├── train.py              # Model training script
├── final_salaries.csv    # Cleaned dataset
├── salary_model.pkl      # Trained model
├── requirements.txt      # Required Python packages
└── README.md             # Documentation file

---

🧪 Screenshots

📌 Dataset Page

View of raw data and structure

📈 Graphs Page

EDA using seaborn and matplotlib

🤖 Prediction Page

Form input and real-time prediction result

---

🌐 Deployment Instructions

1. Push files to GitHub repo
2. Go to: https://streamlit.io/cloud
3. Connect GitHub and select your repo
4. Choose app.py as the main file
5. Click Deploy

---

📚 References

1. 📘 DELNET Research Portal – https://delnet.in
2. 📄 Paper: Salary Prediction Using Machine Learning
3. 📊 Dataset: damalialutfiani/Salary-Dataset-Project
4. 🧰 Libraries: pandas, sklearn, streamlit, joblib, etc.
5. 🌐 Streamlit Cloud: https://streamlit.io/cloud

---

✅ Conclusion

This project built a Salary Prediction model using Random Forest Regression. It provides real-time salary estimates through an interactive web interface, demonstrating how ML can be applied to real-world business problems in HR and recruitment.

---

👤 Author

Prathmesh Gajanan Sose
Email:prathmeshsose93@gmail.com