Foundations of Deep Learning

This unit introduces core concepts, mathematics, and tooling of deep learning, including Python programming, common frameworks, and basic neural architectures (perceptrons, feedforward nets, CNNs, RNNs). Students will learn to preprocess data, implement simple models, and evaluate task-appropriate approaches while considering real-world use and team roles, preparing them for deeper applied units that follow.

Learning Objectives

• Analyze and compare foundational concepts in deep learning and machine learning, including perceptrons, neural networks, convolutional and recurrent architectures, to determine appropriate model types for given problem statements
• Apply Python and at least one deep learning framework to implement, train, and test a simple perceptron and a small neural network for basic classification tasks using prepared datasets
• Demonstrate core mathematical and data-preparation skills (linear combinations, activation functions, loss computation, gradient descent, and basic image/text preprocessing) required to develop and troubleshoot basic neural models
• Evaluate model results and task suitability across classification, generative, and reinforcement scenarios and articulate deployment considerations, collaborative roles, and ethical implications for real-world deep learning projects