Skip to content

An integrated data modeling and model experimentation project, packaged as a Streamlit app for predicting estimated compensation in engineering jobs

Notifications You must be signed in to change notification settings

anopsy/Compensation-prediction

Repository files navigation

Compensation Prediction for Engineering Roles

image

After analyzing the data from the Juniors vs ChatGPT study, I decided to use the compensation estimation data to build an app that predicts compensation for various engineering roles. After cleaning, I built a baseline model and later experimented with different regression models. After hyperparameters tuning I decided to go for the SVC model with rbf kernel. The model was developed using data provided by SourceStack.

Running the App in a Docker Container

To run the Streamlit app in a Docker container, follow these steps:

  1. Clone this repository and open in VS Code.
  2. Install and configure Docker for your operating system. Make sure Docker is running.
  3. Open a terminal or command prompt in the directory of the repo and run the following command:
docker build -t compensation-app .
  1. After the image is built, run a container from the image with the following command:
docker run -p 8501:8501 compensation-app
  1. You can now view your Streamlit app in your browser: http://0.0.0.0:8501

Running the App locally

To run the Streamlit app locally, follow these steps:

  1. Clone this repository and open in VS Code.
  2. Install the required libraries listed in requirements.txt using
pip install -r requirements.txt
  1. Open a terminal or command prompt in the directory of the repo and run the following command:
streamlit run ./comp_app.py
  1. You can now view the Compensation App in your browser by following one of the links from your terminal

Next steps

I would like to fine-tune the model for Data Science jobs specifically.

About

An integrated data modeling and model experimentation project, packaged as a Streamlit app for predicting estimated compensation in engineering jobs

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages