Deep Learning: Getting Started

Course details

As generative AI adoption accelerates, many technical professionals are expected to work with neural networks and AI models without a clear understanding of how they’re built or trained. In this course, instructor Kumaran Ponnambalam provides a practical, hands-on introduction to deep learning, guiding you step by step through designing, training, and evaluating neural network models using Python and modern frameworks. Through real‑world examples in both structured and text data, learn how to build working models while learning the foundational workflows that power today’s AI systems. Along the way, find out how to connect core deep learning concepts to modern generative AI architectures like transformers, embeddings, and foundation models. By the end of this course, you’ll be prepared to confidently evaluate, adapt, and reuse deep learning models in enterprise AI applications.

Learning objectives
Design and train basic deep learning models using Python and Keras.
Evaluate and validate neural network performance by applying loss functions, accuracy metrics, and train/validation/test splits.
Apply deep learning techniques to real datasets, including structured data (Iris classification) and text data (spam detection).
Explain how foundational neural networks connect to modern generative AI systems, such as transformers, embeddings, and foundation models.
Reuse existing architectures and open-source models by evaluating suitability and adapting them to unique use cases.

Concepts

Introduction

Getting started with deep learning
Setting up the environment

Introduction to Deep Learning

What is deep learning
Linear regression
An analogy for deep learning
The perceptron
Artificial neural networks
Training an ANN

Neural Network Architecture

The input layer
Hidden layers
Weights and biases
Activation functions
The output layer

Training a Neural Network

Setup and initialization
Forward propagation
Measuring accuracy and error
Back propagation
Gradient descent
Batches and epochs
An ANN model

Deep Learning - Example 1

Input preprocessing
Creating a deep learning model
Training and evaluation
Saving and loading models
Predictions with deep learning models

Deep Learning and Generative AI

Neural network architectures
Foundation models

Deep Learning - Example 2

Setup for text processing
Creating text representations
Building a spam model
Predictions for text

Deep Learning Exercise

Preprocessing RCA data
Building the RCA model
Predicting root causes with deep learning