Thanks to visit codestin.com
Credit goes to github.com

Skip to content

shayshankr/EEAI-CA

 
 

Repository files navigation

Engineering and Evaluating Artificial Intelligence (EEAI - CA)

Task 1 – Chained Multi-Output Classification

  • Three models are trained independently for Type2, Type3, and Type4 labels.
  • Prediction happens in a chained manner:
    Type2 → Type3 (if Type2 is correct) → Type4 (if both Type2 and Type3 are correct).
  • Final accuracy is calculated based on how many predictions are fully correct across all three stages.
  • Models: 'Random Forest', 'Logistic Regression'.

Task 2 – Hierarchical Classification

  • Prediction is done step-by-step using filtered data at each stage.
  • Model 1 (Type2) predicts → filters data for Model 2 (Type3) → filters again for Model 3 (Type4).
  • Each model is trained only on relevant subsets of data.
  • Models: 'Random Forest', 'Logistic Regression'.
  • Fully modular with runtime model selection per stage.

How to Run

  • Make sure you have Python 3.10+ installed.
  • Install requirements
    • pip install -r requirements.txt
  • Run the main controller
    • python main.py

Dependencies

  • Listed in 'requirements.txt'
  • Main packages include:
    • scikit-learn
    • pandas
    • numpy

About

EEAI-CA (x23305762, x23348186)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%