Deep Learning with Artificial Neural Networks (ANN)

This repository contains implementations of Artificial Neural Networks (ANNs) using TensorFlow/Keras. The models are trained on datasets while experimenting with different activation functions, gradient descent techniques, and optimization algorithms to compare their performance.

🚀 Features

• Built and trained ANN models
• Implemented multiple activation functions
• Compared different optimizers
• Applied data preprocessing and feature scaling
• Evaluated model performance on classification/regression tasks

🛠️ Tech Stack

• Python
• TensorFlow & Keras
• NumPy
• Pandas
• Matplotlib
• Scikit-learn

📚 Topics Covered

• Artificial Neural Networks (ANN)
• Forward & Backpropagation
• Weight Initialization
• Learning Rate, Epochs & Batch Size

Activation Functions

• Sigmoid
• ReLU
• Leaky ReLU
• ELU
• PReLU
• Softmax

Optimizers

• Batch Gradient Descent
• Stochastic Gradient Descent (SGD)
• Mini-Batch Gradient Descent
• Adam
• RMSprop
• Adagrad

📊 Evaluation Metrics

• Accuracy
• Precision
• Recall
• F1 Score
• Confusion Matrix
• MSE / MAE / RMSE
• R² Score

📖 Learning Outcomes

• Build and train ANN models
• Compare activation functions and optimizers
• Understand the impact of hyperparameters on model performance
• Apply ANN to classification and regression problems