Month 5: Deep Learning Basics

Month 5: Deep Learning Basics

Week 1: Introduction to Neural Networks

• Day 1: Introduction to Neural Networks: Perceptrons, Activation Functions
• Day 2: Feedforward Neural Networks: Network Architecture, Weights and Biases
• Day 3: Backpropagation and Gradient Descent
• Day 4: Optimization techniques: Momentum, RMSProp, Adam
• Day 5: Practical application of Neural Networks and a mini-project

Week 2: Convolutional Neural Networks (CNNs)

• Day 1: Introduction to CNNs: Motivation and Basic Concepts
• Day 2: CNN Layers: Convolution, Pooling, Fully Connected
• Day 3: Activation Maps and Feature Visualization
• Day 4: Popular CNN Architectures: LeNet, AlexNet, VGGNet, GoogLeNet, ResNet
• Day 5: Practical application of CNNs and a mini-project (e.g. Image Classification)

Week 3: Recurrent Neural Networks (RNNs)

• Day 1: Introduction to RNNs: Motivation and Basic Concepts
• Day 2: Understanding Hidden States and Vanishing Gradient Problem
• Day 3: Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU)
• Day 4: Bidirectional and Deep RNNs
• Day 5: Practical application of RNNs and a mini-project (e.g. Text Generation)

Week 4: Autoencoders and Generative Adversarial Networks (GANs)

• Day 1: Introduction to Autoencoders: Encoder, Decoder, and Latent Space
• Day 2: Variations of Autoencoders: Denoising, Sparse, and Variational Autoencoders
• Day 3: Introduction to Generative Adversarial Networks (GANs)
• Day 4: GAN Architectures: DCGAN, CycleGAN, StyleGAN