RAIN

LMS login Apply

Cover of AIML 212 Deep Learning

AIML 212

Deep Learning

Neural Networks, Computer Vision, Language Models and Generative AI

Deep learning is the technology behind most of what the public now calls artificial intelligence: phones that recognise faces, apps that diagnose crop diseases from a photograph, systems that transcribe and translate speech, and assistants that write, summarise and answer questions. All of them rest on one idea, developed in this book: networks of simple units, arranged in many layers and trained from data by gradient descent, can learn representations of images, sound and text that no programmer could write by hand.

14 chapters in 4 parts, 294 pages. Editor-in-Chief: Olusola Sayeed Ayoola. Published by RAIN, Ibadan, 2026.

Order a copy Read it (owners and RAIN trainees) Certification

Contents

Open a chapter to see its sections. Each chapter ends with a QR code for its free assessment.

1From Neurons to Networks

1.1 The artificial neuron and the perceptron
1.2 Multilayer perceptrons
1.3 Activation functions
1.4 Universal approximation and the value of depth
1.5 Forward propagation in matrix form
1.6 Implementation

Take the Chapter 1 assessment

2Backpropagation

2.1 Computational graphs
2.2 Backpropagation in a multilayer perceptron
2.3 Vanishing and exploding gradients
2.4 A two-layer network from scratch
2.5 Automatic differentiation

Take the Chapter 2 assessment

3Training Deep Networks

3.1 Loss functions
3.2 Optimisers
3.3 Initialisation and normalisation
3.4 Regularisation
3.5 Workflows: PyTorch, Keras and GPUs
3.6 Experiments

Take the Chapter 3 assessment

4Convolutional Neural Networks

4.1 Convolution
4.2 Padding, stride, pooling and receptive fields
4.3 Implementing and differentiating convolution
4.4 Architectures
4.5 Transfer learning and fine-tuning
4.6 Case study: a cassava disease classifier

Take the Chapter 4 assessment

5Computer Vision Techniques

5.1 Images and OpenCV basics
5.2 Object detection
5.3 Segmentation
5.4 Evaluating detectors
5.5 Case studies

Take the Chapter 5 assessment

6Generative Adversarial Networks and Diffusion Models

6.1 The GAN game
6.2 Why GAN training is hard
6.3 Diffusion models
6.4 Case study: generating from a known distribution

Take the Chapter 6 assessment

7Recurrent Networks

7.1 Recurrent neural networks
7.2 Backpropagation through time
7.3 LSTM and GRU
7.4 Sequence-to-sequence models
7.5 Case studies

Take the Chapter 7 assessment

8Natural Language Processing

8.1 Text preprocessing and tokenisation
8.2 Bag-of-words, TF-IDF and text classification
8.3 Word embeddings
8.4 Sequence labelling and named-entity recognition
8.5 NLP for Nigerian languages
8.6 Case studies

Take the Chapter 8 assessment

9Transformers and Attention

9.1 Attention as a differentiable lookup
9.2 Multi-head attention, masking and the transformer block
9.3 Positional encoding
9.4 Transformer architectures
9.5 Complexity and efficient attention
9.6 Case studies

Take the Chapter 9 assessment

10Large Language Models

10.1 Pretraining
10.2 Decoding
10.3 Fine-tuning
10.4 Instruction tuning and alignment
10.5 Retrieval-augmented generation
10.6 Tools and agents
10.7 Evaluating language-model systems
10.8 Case studies

Take the Chapter 10 assessment

11Explainable AI

11.1 What is an explanation?
11.2 LIME
11.3 Shapley values
11.4 Gradient-based attributions
11.5 The limits of explanation
11.6 Case studies

Take the Chapter 11 assessment

12Generative AI Applications

12.1 Text generation and prompting
12.2 Image generation
12.3 Audio and speech
12.4 Watermarking and provenance
12.5 Copyright, consent and misuse
12.6 Case studies

Take the Chapter 12 assessment

13Deploying Deep Learning Models

13.1 Model formats
13.2 Serving models
13.3 Models on phones and small devices
13.4 Monitoring and drift
13.5 Deploying in Nigerian conditions
13.6 Case studies

Take the Chapter 13 assessment

14Responsible AI

14.1 Fairness
14.2 Privacy
14.3 Safety and robustness
14.4 Governance and documentation
14.5 Case studies

Take the Chapter 14 assessment

Order a copy

Printed copies are delivered anywhere in Nigeria or collected at RAIN in Ibadan; digital copies arrive by e-mail with your book owner code. We will reply with the price, delivery cost and payment details.

Schools and companies ordering several copies: put the number you need and we will quote.

Other RAIN textbooks

AIML 101 Building and Deploying AI Applications

AIML 111 Data Science with Python

AIML 112 Machine Learning I

AIML 211 Machine Learning II

RDA 111 Product Design and Development

RDA 112 Practical Electronics