This folder is a dedicated repo-style package for Problem 2: Logistic Regression.
This repository is part of the ML Arena family, a set of 5 dedicated repositories where each repository focuses on one core ML problem.
- ML-Arena-1 - Linear Regression
- ML-Arena-2 - Logistic Regression (Current Repo)
- ML-Arena-3 - Decision Tree
- ML-Arena-4 - K-Means Clustering
- ML-Arena-5 - Neural Network
- File:
dataset.csv - Source: Pima Indians Diabetes Dataset (UCI)
- Rows: 768
- Columns: 9 (8 input features + 1 label)
- Target:
Outcome
Logistic-Regression/
├─ README.md
├─ CONTRIBUTING.md
├─ dataset.csv
├─ exploration/
│ └─ exploration.ipynb
├─ library/
│ └─ training.ipynb
├─ scratch/
│ └─ training.ipynb
└─ PULL_REQUEST_TEMPLATE.md
- Exploration: Class balance, feature behavior, and data quality.
- Library: Baseline logistic model with standard libraries.
- Scratch: Sigmoid + loss + gradient descent using NumPy.
- Optimization: Improve baseline using threshold tuning, regularization, or class weighting.
Each issue is open to multiple contributors.
- Multiple PRs can be submitted for the same issue.
- Different approaches, implementations, and optimizations are welcome.
This means even if someone has already submitted a PR for an issue, you are still encouraged to submit your own solution.
- Fork this folder as its own repo (Logistic-Regression).
- Clone your fork and create a branch.
- Pick one issue.
- Work in only one notebook based on track:
exploration/exploration.ipynblibrary/training.ipynbscratch/training.ipynb
- Run all cells top to bottom.
- Open a pull request with
Closes #<issue-number>.
See CONTRIBUTING.md for detailed rules.