https://github.com/rihua-tech/iris-ml-end-to-end
End-to-end Iris classification in Python: EDA → stratified CV → model comparison → SVM grid search → hold-out test → model persistence.
https://github.com/rihua-tech/iris-ml-end-to-end
classification eda machine-learning python scikit-learn svm
Last synced: about 1 month ago
JSON representation
End-to-end Iris classification in Python: EDA → stratified CV → model comparison → SVM grid search → hold-out test → model persistence.
- Host: GitHub
- URL: https://github.com/rihua-tech/iris-ml-end-to-end
- Owner: rihua-tech
- Created: 2025-09-25T00:43:32.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2025-09-25T01:40:05.000Z (9 months ago)
- Last Synced: 2025-09-25T03:31:34.348Z (9 months ago)
- Topics: classification, eda, machine-learning, python, scikit-learn, svm
- Language: Python
- Homepage:
- Size: 10.8 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.txt
Awesome Lists containing this project
README
ITEC 6720 — Final Project (README)
WHAT THIS IS
- Lesson 1: Mini-course snippets (small standalone scripts).
- Lesson 2: Step-by-Step Iris project (full ML workflow: load → EDA → CV → tune SVM → test → save model).
FOLDERS
- scripts/mini_course/ – MC-01 … MC-14 scripts
- scripts/step_by_step/ – 01_env_versions.py … 06_predictions.py, plus
- iris_end_to_end_SVM_model.py (main)
- iris_end_to_end_NB_model.py (optional)
- data/ (optional local iris.csv)
- figures/, models/, reports/
ENV / INSTALL
Python 3.x
pip install numpy scipy pandas matplotlib scikit-learn joblib
HOW TO RUN (Step-by-Step)
python scripts/step_by_step/iris_end_to_end_SVM_model.py
This script: loads data, makes EDA plots, runs 10-fold CV spot-check, grid-searches SVM, evaluates on test, saves model and summary.
(Optional) Auto-select variant:
python scripts/step_by_step/iris_end_to_end_NB_model.py
KEY OUTPUTS (generated)
- Figures: figures/eda_boxplots.png, eda_hist.png, eda_scatter_matrix.png,
algo_boxplot.png, cv_vs_test.png, confusion_matrix.png
- Reports: reports/spotcheck_cv.csv, step_by_step_run_summary.txt
- Model: models/iris_best_model.joblib
- Console shows: 10-fold CV scores, best SVM params, TEST accuracy (~0.92), confusion matrix, classification report.
SUBMIT CHECKLIST
[ ] Scripts (mini_course + step_by_step)
[ ] Screenshots doc with captions
[ ] 500-word report (a–g)
[ ] Generated figures/reports/model
[ ] This README.txt