Iris Flower Classification using K-Nearest Neighbors (KNN)

Internship Task 6 – Distance-Based ML Model
Submitted by: Tirtha Dutta
Model Type: Multiclass Classification | Supervised Learning

---

Objective

Build and evaluate a K-Nearest Neighbors (KNN) model to classify species of iris flowers based on petal and sepal measurements. Demonstrate complete ML workflow from data cleaning to visual decision boundaries.

---

Dataset

• Source: Iris Dataset – Kaggle
• Classes: Iris-setosa, Iris-versicolor, Iris-virginica
• Shape (after cleaning): 150 rows × 5 columns
• Target Variable: Species

---

Workflow Summary

1️ Data Cleaning

• Dropped unnecessary Id column
• Verified no missing/null values
• Saved cleaned version as iris_cleaned.csv

2️ Label Encoding & Feature Scaling

• Encoded flower names (Species) to numeric labels: 0, 1, 2
• Standardized all features using StandardScaler

3️ Exploratory Data Analysis (EDA)

• Generated boxplots to visualize feature distributions
• Created a correlation heatmap
• Analyzed class distribution, feature types, and outliers

4️ Model Building: KNN Classifier

• Explored different values of K (1 to 15)
• Best accuracy at K = 1: 96.67%
• Trained final KNN model on scaled data

5️ Model Evaluation

• Evaluated using accuracy, confusion matrix, and classification report
• Saved confusion matrix proof as: images/confusion_matrix.png

6️ Visual Decision Boundary

• Trained a 2D KNN model on first two features
• Created a mesh grid to plot class separation
• Saved output as: images/knn_decision_boundary.png

---

Final Results

Metric	Value
Accuracy (K = 1)	96.67%
Number of Classes	3
Best K Value	1

---

Tools & Libraries Used

• Python
• pandas, numpy
• scikit-learn
• matplotlib, seaborn
• Jupyter Notebook

---

Skills Demonstrated

• Data Cleaning & Preprocessing (Task 1)
• Exploratory Data Analysis (Task 2)
• Feature Scaling & Label Encoding (Task 3)
• Classification using KNN (Task 4 & Task 6)
• Confusion Matrix & Evaluation Reports
• Visual Decision Boundary Plot (Task 6B)
• GitHub project structuring & documentation

---

This project is built for internship evaluation and job-readiness. All steps are implemented with professional standards, including proofs from Tasks 1–6.