This project aims to develop a Machine Learning pipeline to predict income levels based on the Adult Income Census dataset. The pipeline includes various components such as a custom logging system, custom exception handling and deliverables ensuring robust and transparent operation throughout the ML workflow.
Becker, B. & Kohavi, R. (1996). Adult [Dataset]. UCI Machine Learning Repository. https://doi.org/10.24432/C5XW20.
- Data Ingestion: Collecting and loading the dataset for processing.
- Exploratory Data Analysis (EDA): Analyzing the dataset to identify patterns, distributions, and correlations.
- Data Transformation: Preprocessing and transforming the data for model readiness.
- Model Training: Training various models to predict income levels.
- Model Evaluation: Evaluating model performance using appropriate metrics.
- Model Deployment: Deploying the model to a Flask server for making predictions on new data.
- Logger and Exception Handler: Custom logging and exception handling mechanisms are integrated to provide transparency, easy debugging, and robust error management across the pipeline.
- Artifacts/Deliverables: All key artifacts such as preprocessed data transformations, trained models, and important EDA insights are saved and versioned in the artifacts/ directory for reproducibility and further analysis. The project demonstrates end-to-end model development and deployment while leveraging custom logging and error handling for a production-ready solution.
Install Anaconda to manage project dependencies.
Open Anaconda Terminal, create a new envrionment with python 3.10 installed.
conda create -p env/ python=3.10
Navigate to folder where env is located. Activate conda envrionment.
conda activate env/
You should see the path of the enviroment on the leftmost hand side, indicating successfull activation.
Clone the project
git clone https://github.com/Tryd3x/ml-pipeline.git
Go to the project directory
cd ml-pipeline
Activate conda environment
conda activate env/
Install dependencies using pip
pip install -r "requirements.txt"
Start the Flask server
python app.py
If you have any feedback, please reach out to us at htelegraphy@gmail.com