🤖 Machine Learning Algorithms Collection

Where classical statistics meets modern AI wizardry

"In the garden of algorithms, every flower tells a different story of classification" 🌸

📚 Overview

This repository contains a comprehensive collection of supervised machine learning algorithms implemented during the AI course at IFSULDEMINAS – Campus Muzambinho. Each algorithm is meticulously crafted and tested on the legendary Iris dataset – because sometimes the classics are classics for a reason.

Think of this as your digital herbarium of machine learning models, where each algorithm is a different lens through which we can understand the art of classification.

🔬 Algorithms Implemented

🎯 Logistic Regression (regressao_logistica.ipynb)

The elegant statistician of the bunch. Despite its name suggesting regression, it's actually a classification powerhouse that uses the sigmoid function to squeeze probabilities into neat little boxes.

Key Features: - Binary and multiclass classification - Probabilistic outputs - Linear decision boundaries - Interpretable coefficients

🎲 Naive Bayes (Naive_Bayes.ipynb)

The optimistic probabilist who assumes independence even when the world is beautifully interconnected. Sometimes naive assumptions lead to surprisingly sophisticated results.

Key Features: - Bayesian probability framework - Fast training and prediction - Handles categorical and continuous features - Excellent for text classification (though we're using flowers here)

🌳 Random Forest (Random_Forest.ipynb)

Democracy in action – where multiple decision trees vote to reach a consensus. Because wisdom of crowds often trumps individual genius.

Key Features: - Ensemble of decision trees - Built-in feature importance - Handles overfitting gracefully - Robust to outliers

⚡ XGBoost (Treinado_com_XGBoost.ipynb)

The speed demon with a PhD in gradient optimization. When you need results yesterday and accuracy tomorrow.

Key Features: - Extreme gradient boosting - High performance and scalability - Advanced regularization - Feature importance ranking

💡 LightGBM (Gradient_Boosting_usando_LightGBM.ipynb)

The minimalist's dream – maximum efficiency with minimum memory footprint. Light on resources, heavy on results.

Key Features: - Leaf-wise tree growth - Memory efficient - GPU acceleration support - Fast training speed

🐱 CatBoost (Gradient_Boosting_CatBoost.ipynb)

The categorical data whisperer. Handles mixed data types like a Swiss Army knife handles camping situations.

Key Features: - Native categorical feature support - Symmetric trees - Ordered boosting - Minimal hyperparameter tuning

🌸 Dataset: The Timeless Iris

The Iris dataset – Ronald Fisher's 1936 gift to machine learning that keeps on giving. Three species of iris flowers (setosa, versicolor, virginica) measured across four features:

• Sepal Length (cm)
• Sepal Width (cm)
  
• Petal Length (cm)
• Petal Width (cm)

150 samples, 3 classes, infinite possibilities.

🚀 Getting Started

Prerequisites

```bash

The usual suspects

pip install numpy pandas scikit-learn matplotlib seaborn pip install xgboost lightgbm catboost pip install jupyter notebook ```

Quick Start

```bash

Clone this digital garden

git clone https://github.com/anderson-ufrj/machine-learning-algorithms.git cd machine-learning-algorithms

Fire up Jupyter

jupyter notebook

Pick your algorithm and let the magic begin ✨

```

📊 Performance Metrics

Each notebook includes comprehensive evaluation metrics:

• Accuracy Score - The classic crowd-pleaser
• Precision & Recall - For when you need to be more nuanced
• F1-Score - The harmonic mean that keeps everyone happy
• Confusion Matrix - Visual truth in a grid
• Classification Report - The full statistical symphony
• Cross-Validation - Because one split is never enough

🎨 Visualizations

Every notebook comes with rich visualizations: - Decision boundaries (where possible) - Feature importance plots - Learning curves - Confusion matrices with heatmaps - ROC curves for the probability enthusiasts

🏗️ Project Structure

├── 📁 notebooks/ │ ├── 📓 regressao_logistica.ipynb │ ├── 📓 Naive_Bayes.ipynb │ ├── 📓 Random_Forest.ipynb │ ├── 📓 Treinado_com_XGBoost.ipynb │ ├── 📓 Gradient_Boosting_usando_LightGBM.ipynb │ └── 📓 Gradient_Boosting_CatBoost.ipynb ├── 📄 README.md └── 📄 requirements.txt

🎯 Learning Objectives

This collection serves multiple pedagogical purposes:

1. Algorithm Understanding - Deep dive into how each method thinks
2. Comparative Analysis - See how different approaches tackle the same problem
   
3. Implementation Skills - Hands-on coding with real libraries
4. Evaluation Techniques - Master the art of model assessment
5. Visualization - Tell stories with data and graphs

🔮 Future Enhancements

• [ ] Deep learning implementations (Neural Networks)
• [ ] Unsupervised learning algorithms
• [ ] Advanced ensemble methods
• [ ] Hyperparameter optimization tutorials
• [ ] Model deployment examples
• [ ] Performance benchmarking suite

📖 References & Inspiration

• Fisher, R.A. (1936). "The use of multiple measurements in taxonomic problems"
• Scikit-learn documentation
• XGBoost, LightGBM, and CatBoost official documentation
• IFSULDEMINAS AI Course Materials

🤝 Contributing

Found a bug? Have an idea? Want to add another algorithm to the collection?

```bash

Fork, code, commit, push, PR

The eternal dance of open source collaboration

```

📜 License

MIT License – because knowledge should be free to roam and reproduce.

👨‍💻 Author

Anderson Silva - Digital Intelligence Architect - 🔗 LinkedIn: anderson-h-silva95 - 🐦 Twitter: @neural_thinker - 📧 Email: andersonhs27@gmail.com - 🐙 GitHub: anderson-ufrj

🎵 Coda

"In the symphony of machine learning, each algorithm plays its own instrument. Some are violins – precise and melodic. Others are drums – powerful and rhythmic. Together, they create the music of artificial intelligence."

---

Tags: #MachineLearning #AI #Python #Scikit-learn #DataScience #Classification #Algorithms #IFSULDEMINAS #Iris #XGBoost #LightGBM #CatBoost #RandomForest #NaiveBayes #LogisticRegression

Made with ❤️ and countless cups of coffee in Muzambinho, MG ☕