Syllabus
Section 1: Introduction to Deep Learning: Overview of deep learning and its significance in artificial intelligence, History and evolution of deep learning, Key applications and use cases of deep learning in geospatial and remote sensing domain
Section 2: Feed-forward Neural Networks, Introduction to neural networks and their basic components
Activation functions and their role in deep learning, Forward pass and prediction in a neural network
Loss functions and gradient descent for model optimization, Training a feed-forward neural network
Case Study Problem Introduction: Using simple neural networks
Section 3: Backpropagation, Understanding the concept of backpropagation, Calculating gradients and updating weights, Optimization algorithms (e.g., stochastic gradient descent), Regularization techniques to prevent overfitting, Case Study Solution: Using simple neural network in geospatial analysis
Section 4: Recurrent Neural Networks (RNNs), Introduction to RNNs and their architecture, Vanishing and exploding gradients problem,Long Short-Term Memory (LSTM) units, Training and applying RNNs for sequential data processing tasks, Case Study: Problem solving using Recurrent Neural Networks
Section 5: Attention Mechanisms, The need for attention mechanisms in deep learning, Self-attention and transformer architecture Applications of attention mechanisms in natural language processing and computer vision tasks. Case Study: Problem solving using Attention models
Section 6: Transformers, Introduction to transformer architecture and its components, Multi-head attention and positional encoding Transformer-based models for machine translation and language understanding Case Study Solution: Problem solving using Transformers