Project overview

SalarySense AI

SalarySense AI is a platform that provides valuable insights into salary trends across various job roles, countries, and industries. It leverages machine learning algorithms and data analysis to deliver up-to-date salary predictions and detailed visualizations for professionals and employers alike.

Table of Contents

• Introduction
• Features
• Installation
• Usage
• Modeling Process
• Dashboard and Visualization
• Technologies
• Contributing
• License

Introduction

SalarySense AI aims to provide an intuitive, data-driven platform for understanding salary trends across various job markets and demographics. The project combines data analysis, machine learning, and visualization tools to offer a holistic view of salary distribution, predictions, and related insights.

Features

• Salary Prediction Model: Predicts salary based on factors such as job title, location, years of experience, education level, and more.
• Data Visualization: Provides visual insights using Tableau and Plotly dashboards to show trends in salary distribution by country, year, and job type.
• Exploratory Analysis: Allows users to explore salary data and see relationships between job characteristics and salary trends.
• Streamlit App: An interactive user interface to predict salaries and explore trends.

Installation

1. Clone the repository:

git clone https://github.com/YourUsername/SalarySenseAI.git

2. Install dependencies:

pip install -r requirements.txt

3. Run the Streamlit App:

streamlit run app/app.py

4. Run the Flask Backend (if applicable):

python app/flask_app.py

Usage

The project consists of two key components:

1. Salary Prediction: Users can input specific details (job title, country, experience, etc.) to get a salary prediction based on the machine learning model.
2. Exploratory Analysis: Users can visualize and interact with salary trends using interactive charts, either through Streamlit or Tableau dashboards.

Modeling Process

1. Data Collection: The dataset used for the model consists of developer salary data from 2021 to 2024, collected from Stack Overflow yearly developer survey.
2. Data Cleaning and Preprocessing: Handling missing data, transformed columns like job titles and locations using label encoding, and standardized the salary data to EUR for uniformity.
3. Feature Engineering: Relevant features such as years of experience, education, country, and job type were selected.
4. Model Selection: A variety of models were tried, including Linear Regression, Decision Trees, and Random Forests. The Random Forest model was fine-tuned using hyperparameter tuning and cross-validation.
5. Model Evaluation: The Random Forest model performed the best, with an R² of 0.55, which was selected for the final prediction.

Dashboard and Visualization

1. Streamlit App:
   • Prediction Page: Allows users to predict salaries based on selected criteria.
   • Explore Page: Users can visualize salary distributions across various countries, job types, and demographics.
2. Tableau Dashboard: A comprehensive dashboard that shows salary trends across different job roles and countries for further exploration.

Technologies

• Python: Core programming language used for data analysis, model building, and backend development.
• Pandas and NumPy: Used for data manipulation and analysis.
• Scikit-learn: For building and evaluating machine learning models.
• Streamlit: For building an interactive frontend application.
• Tableau: Used for creating rich, interactive visualizations.
• Flask: Backend framework for serving web pages and APIs.
• Plotly: For generating interactive visualizations.
• Pickle: For model serialization and saving encoders.