Neural Networks and Deep Learning

5-6 weeks

LearningPerceptionNatural Interaction

Overview

This unit introduces neural networks, the foundation of modern deep learning. Students will understand how neural networks learn through backpropagation, build networks using Keras, and explore specialized architectures like CNNs for image recognition. They will grapple with training challenges (overfitting, vanishing gradients) and learn practical solutions (regularization, early stopping). This unit demystifies deep learning and empowers students to build and reason about neural network systems.

Learning Objectives

1
Understand the structure and function of artificial neurons and how networks of neurons learn.
2
Implement and train neural networks using Keras/TensorFlow.
3
Understand convolutional neural networks (CNNs) and their application to image recognition.
4
Diagnose and address overfitting through regularization, dropout, and early stopping.
5
Apply transfer learning to leverage pre-trained models.

Sign in to access full content

Create a free account to access lesson plans, projects, assessments, and downloadable resources for this unit.