Salary Prediction: a Data Science Sample Project

About

This platform showcases a supervised learning project aimed at predicting workers’ salaries using various data science techniques.

Project Overview

The project involves a comprehensive analysis of factors influencing salary levels, employing data cleaning, exploratory data analysis, and the development of predictive models. The goal is to accurately forecast salaries based on relevant features.

Website Structure

• Insights: Dive into data exploration, including data cleaning and exploratory analysis, to understand the dataset’s intricacies.
• Models: Explore the modeling process, featuring train/test data splitting, implementation of linear regression and large language models, and a comparison of their performances.
• Predictions for New Data: Apply the trained models to new data inputs to predict potential salary outcomes.

Good Practices in Project Development

This project follows good practices to ensure reproducibility and maintainability:

• Reproducibility: All steps are documented and structured to be easily replicable.
• Version Control: The entire project is managed using Git and hosted on GitHub.
• Environment Management: A Conda environment is provided to guarantee dependency consistency.
• Quarto for Documentation: Instead of traditional Jupyter notebooks, Quarto has been used for better organization and presentation of results.
• Ease of making new predictions: through an app developed with Shiny for Python, hosted in this repo.