Machine Learning Notebooks - Focussed on CNNs 📓🤖

Welcome to this repository of Jupyter Notebooks where I explore and document key concepts, techniques, and models in Machine Learning and Artificial Intelligence. Each notebook is designed to be beginner-friendly yet informative, covering both theory and implementation.

📁 Current Notebooks

1. Foundations of Machine Learning

• foundations.ipynb: A practical walkthrough explaining how machine learning models are trained, including data splitting, model fitting, loss calculation, and optimization.

2. Computer Vision Basics

• 0 neural networks.ipynb: Introduction to neural networks
• 1 model.ipynb: Building a neural network to classify images of clothing using Fashion MNIST
• 2 callback model.ipynb: Implementation of callbacks for model training optimization
• 3 updating model.ipynb: Exploring model architecture modifications and their impacts

3. Convolutional Neural Networks

• convolutions and pooling.ipynb: Deep dive into CNN fundamentals

4. Training in Cloud vs Local

• model to run in vertex ai.ipynb: Training a CNN model in Google Cloud Vertex AI
• model to run locally.ipynb: Local implementation of the same CNN model

5. Complex Image Classification

• horse or human - local.ipynb: Training a CNN to distinguish between horses and humans (1283 images, 300x300 resolution)
• horse or human - vertex ai.ipynb: Cloud-based implementation of the horse/human classifier

🧮 Key Topics Covered

• Neural Network Fundamentals
• Computer Vision & Image Classification
• Convolutional Neural Networks (CNNs)
• Model Training and Evaluation
• Cloud vs Local Development
• Callbacks and Model Optimization
• Data Preprocessing and Normalization
• Model Architecture Design
• Complex Image Classification

🛠 Requirements

To run the notebooks locally, you need:

• Python 3.10 or less because of TensorFlow compatibility
• Jupyter Notebook or JupyterLab
• Key libraries:

  pip install tensorflow tensorflow-datasets numpy matplotlib
  pip install google-cloud-aiplatform  # For Vertex AI notebooks

🏗 Project Structure

• Each section is organized in numbered folders
• Includes both local and cloud (Vertex AI) implementations
• Supporting files (images, models) are included where needed
• Clear progression from basic concepts to complex applications