Iris Flower Classifier

An end-to-end machine learning pipeline for classifying Iris species. From data preparation and model training to evaluation and a Flask API for live predictions.

---

🚀 Features

• Data Preparation
  • Load raw CSV, minimal cleaning, stratified train/test split
• Model Training
  • Standard scaling + Logistic Regression
• Evaluation
  • Accuracy, classification report, confusion matrix
• API Service
  • Flask app exposing /predict endpoint for JSON input
• Serialization
  • Model saved as models/iris_classifier.joblib

---

📁 Repository Structure

iris-classifier/ ├── data/ │ ├── raw/ # raw iris.csv │ └── processed/ # train.csv & test.csv │ ├── images/ │ └── iris_flowers.png # illustration used in README │ ├── models/ │ └── iris_classifier.joblib # trained model pipeline │ ├── src/ │ ├── data_prep.py # data loading & train/test split │ ├── train.py # build & train model pipeline │ ├── evaluate.py # compute metrics & print reports │ └── app.py # Flask API for live inference │ ├── requirements.txt # project dependencies └── README.md # this file

---

⚙️ Setup & Installation

1. Install dependencies

pip install -r requirements.txt

2. Prepare the data

python src/data_prep.py

3. Train the model

python src/train.py
--input data/processed/train.csv
--output models/iris_classifier.joblib

4. Evaluate Performance

python src/evaluate.py
--model-path models/iris_classifier.joblib
--test-data data/processed/test.csv

🖥️ Running the Flask API

cd src flask run --host=0.0.0.0 --port=5000

/predict Endpoint

Send a POST request with JSON body: { "sepal_length": 5.1, "sepal_width": 3.5, "petal_length": 1.4, "petal_width": 0.2 }

Response

{ "predicted_species": "Setosa", "confidence": 0.96 }