Skip to content

Latest commit

 

History

12 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MLOps Experiment Tracking with Git

A coursework project demonstrating manual machine learning experiment tracking and Git-based version control practices using Decision Tree experiments on the Palmer Penguins dataset.

This project was completed as part of DSC524 – Designing MLOps for Enterprises in the M.Tech Artificial Intelligence and Data Science program at IIIT Kottayam.


🎯 Project Objective

The objective of this project was to understand foundational MLOps practices by tracking changes across multiple machine learning experiments.

The project focuses on:

  • Git-based version control
  • Experiment tracking
  • Hyperparameter experimentation
  • Preprocessing variations
  • Feature-selection variations
  • Model evaluation
  • Comparing experiment results

🐧 Dataset

The experiments use the Palmer Penguins dataset, containing observations of three penguin species:

  • Adelie
  • Chinstrap
  • Gentoo

The dataset includes physical and categorical characteristics such as bill measurements, flipper length, body mass, island, and sex.

Detailed dataset information is available in dataset_info.md.


🧪 Experiment Tracking

Four Decision Tree experiments were manually tracked with variations in:

  • Hyperparameters
  • Preprocessing
  • Feature selection
  • Precision
  • ROC-AUC

The best recorded experiment achieved:

  • Precision: 0.99
  • ROC-AUC: 0.99

The complete experiment table is available in:

experiment_tracking.md


📊 Experiment Summary

Experiment Configuration Precision ROC-AUC
EXP-01 Baseline Decision Tree 0.97 0.97
EXP-02 Max Depth = 5 + preprocessing 0.98 0.98
EXP-03 Max Depth = 7 + preprocessing + imputation 0.99 0.99
EXP-04 Pruned Decision Tree 0.98 0.99

🔄 MLOps Workflow Demonstrated

Baseline Experiment
        ↓
Record Configuration & Metrics
        ↓
Modify Hyperparameters / Preprocessing
        ↓
Run New Experiment
        ↓
Compare Results
        ↓
Track Changes with Git
        ↓
Document Findings

📂 Repository Structure

mlops-git-experiment-tracking/
├── README.md
├── analysis.md
├── dataset_info.md
├── experiment_tracking.md
└── .gitignore

Files

  • README.md — Project overview and experiment summary
  • analysis.md — Findings from the recorded experiments
  • dataset_info.md — Palmer Penguins dataset information
  • experiment_tracking.md — Manual experiment tracking table

💡 What This Project Demonstrates

This project demonstrates foundational MLOps concepts including:

  • Tracking ML experiments systematically
  • Comparing model configurations
  • Recording evaluation metrics
  • Managing experiment changes using Git
  • Documenting model-development decisions

⚠️ Project Scope

This repository focuses on Git-based workflow and manual experiment tracking.

The model-training implementation is not included in this repository. More advanced experiment tracking using tools such as MLflow is covered in later MLOps projects.


👤 Author

Anupam K Ajith AI/ML & GenAI | M.Tech AI & Data Science @ IIIT Kottayam | Python | RAG | MLOps

About

Manual ML experiment tracking and Git-based MLOps workflow using Decision Tree experiments on the Palmer Penguins dataset.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors