Classify iris flowers into three species (Setosa, Versicolor, Virginica) based on measurements of their petals and sepals.

The classic Iris dataset from the UCI Repository, loaded via scikit-learn:

• 150 samples (50 per species)
• 4 features:
  • Sepal length (cm)
  • Sepal width (cm)
  • Petal length (cm)
  • Petal width (cm)
• 3 classes: Setosa, Versicolor, Virginica

• Python 3.x
• Libraries:
  • pandas - Data manipulation
  • numpy - Numerical operations
  • matplotlib - Visualization
  • seaborn - Advanced visualization
  • scikit-learn - Machine learning models and metrics

• Load the Iris dataset from scikit-learn
• Create a pandas DataFrame for easier manipulation
• Display basic information and statistics

• Pairplot: Visualize relationships between all feature pairs
• Histograms: Show distribution of each feature by species
• Box Plots: Display feature distributions and outliers
• Correlation Heatmap: Show correlation between features

• Check for missing values (none found)
• Split data into training (80%) and test (20%) sets
• Apply feature scaling using StandardScaler
• Use stratified sampling to maintain class balance

Train and compare three classifiers:

• Logistic Regression
• K-Nearest Neighbors (KNN)
• Decision Tree Classifier

Evaluate models using multiple metrics:

• Accuracy: Overall correctness
• Precision: Positive prediction accuracy
• Recall: True positive detection rate
• F1-Score: Harmonic mean of precision and recall
• Confusion Matrix: Detailed prediction breakdown

Analyze which features are most important for classification using the Decision Tree model.

Install required libraries:

pip install pandas numpy matplotlib seaborn scikit-learn

Run the main script:

python iris_classification.py

All three models typically achieve high accuracy (95%+) on this dataset:

• Logistic Regression: ~97-100%
• K-Nearest Neighbors: ~97-100%
• Decision Tree: ~97-100%

The script generates the following visualization files:

1. iris_pairplot.png - Scatter plots of all feature combinations
2. iris_distributions.png - Histograms showing feature distributions
3. iris_boxplots.png - Box plots for each feature by species
4. iris_correlation_heatmap.png - Feature correlation matrix
5. iris_confusion_matrices.png - Confusion matrices for all models
6. iris_model_comparison.png - Performance metrics comparison
7. iris_feature_importance.png - Feature importance ranking

• ✅ Loading and exploring datasets
• ✅ Data visualization techniques (scatter plots, histograms, heatmaps)
• ✅ Data preprocessing and scaling
• ✅ Train-test split methodology
• ✅ Classification modeling with multiple algorithms
• ✅ Model evaluation using various metrics
• ✅ Confusion matrix interpretation
• ✅ Feature importance analysis
• ✅ Model comparison and selection

1. Petal measurements (length and width) are typically more discriminative than sepal measurements
2. Setosa is linearly separable from the other two species
3. Versicolor and Virginica have some overlap, making them slightly harder to distinguish
4. All three simple classifiers perform excellently on this dataset
5. The dataset is well-balanced with no missing values

• UCI Machine Learning Repository - Iris Dataset
• Scikit-learn Documentation
• Original Paper by R.A. Fisher (1936)

Created as a beginner-friendly machine learning project to demonstrate classification techniques.

This project is open source and available for educational purposes.