ML‑Algorithms

A personal portfolio project showcasing Python proficiency through from‑scratch implementations of classical machine‑learning algorithms. Only NumPy and Matplotlib are used for numerical work and visualisation.

---

Notebooks at a Glance

Algorithm	Notebook	What it demonstrates
Linear Regression	Linear_Regression.ipynb	Closed‑form normal equation, RIDGE and LASSO regression, batch gradient descent, synthetic data fit
Logistic Regression	Logistic_Regression.ipynb	Binary classification, cross‑entropy loss, decision boundary visualisation
k‑Nearest Neighbours	kNN.ipynb	Lazy learning, distance metrics, effect of k on accuracy
Feed‑forward Neural Network	Deep Neural Networks.ipynb	Multi‑layer perceptron built from scratch, back‑prop, mini‑batch SGD

(Each notebook is fully self‑contained; no external datasets or helper libraries are required.)

---

Running the Notebooks

# 1. Clone the project
git clone https://github.com/M6LI/ML-Algorithms.git
cd ML-Algorithms

# 2. (Optional) create a virtual environment
python -m venv .venv && source .venv/bin/activate

# 3. Install minimal dependencies
pip install numpy matplotlib jupyter

# 4. Fire up Jupyter
jupyter notebook

Run the cells top‑to‑bottom to see the algorithms learn in real‑time.

---

Repository Layout

ML-Algorithms/
├── Deep Neural Networks.ipynb
├── Linear_Regression.ipynb
├── Logistic_Regression.ipynb
├── kNN.ipynb
└── README.md

---

License

Released under the MIT License.

---

Built for learning and demonstration!