Artificial Intelligence Expert: Mastering Advanced Techniques - IABAC

Module 1: Mastering the Foundations of Artificial Intelligence

• The concept of intelligence, from human to machine
• History and evolution of artificial intelligence
• Reasons behind the current rise of AI
• Different domains of AI applications
• Distinction between AI, Data Science, and Machine Learning.

Module 2: Mastering the Basics of Machine Learning and Deep Learning

• Fundamental differences between Machine Learning and Deep Learning
• Architecture and functioning of deep neural networks
• Feature learning in deep networks
• Real-world applications of Deep Learning networks

Module 3: Getting Started with TensorFlow

• Overview of the open-source tool TensorFlow
• TensorFlow structure and modules
• Tensors, operations, graphs, variables, and functions
• Building and training simple models
• Dynamic execution (Eager Execution) and compilation (XLA)

Hands-on Lab:

• Building a Machine Learning model with TensorFlow using the Keras API

Module 4: Studying Artificial Neural Networks (ANNs)

• Structure and organization of neural networks
• Core concepts (weight initialization, optimizers, activation, MSE and RMSE)
• Forward propagation algorithm
• Gradient backpropagation for learning

Module 5: Exploring Computer Vision

• Introduction to computer vision and its applications
• Basics of image processing and related techniques
• Convolutional Neural Networks (CNNs): architecture and functioning
• Object detection (concepts, methods, and metrics)
• Bounding Box Regression for localization
• Advanced object detection models (R-CNN, Fast R-CNN, Faster R-CNN, SSD, and YOLO)
• Implementation with OpenCV (CV2) for practical applications
• Transfer learning in CNNs
• Image classification with the Flowers dataset (TF 2.x)

Case Study:

• Analyzing anomaly detection in medical X-rays

Hands-on Lab:

• Image classification with a CNN (e.g., cats vs. dogs)

Module 6: Mastering Natural Language Processing (NLP)

• Introduction to NLP and its challenges
• Core NLP concepts (tokenization, lemmatization, and syntactic analysis)
• Bag of Words (BoW) models for text representation
• Word embeddings and vector word representations
• Handling and processing text and PDF files
• Advanced regex techniques
• Using Transformers and BERT for NLP
• Using GPT models for text generation
• State-of-the-art techniques and advanced NLP projects

Hands-on Lab:

• Applying the BERT algorithm to text processing

Module 7: Discovering Reinforcement Learning (RL)

• Markov Decision Processes (MDP) for RL
• Core RL equations (Bellman equations and key concepts)
• Differences between model-based and model-free RL
• Dynamic programming and model-free methods

Module 8: Advancing into Deep Reinforcement Learning (Deep RL)

• Deep Q-Learning architectures
• Deep Q-Learning and advanced algorithms
• Reinforcement learning projects with OpenAI Gym

Module 9: Mastering Prompt Engineering and Interacting with LLMs

• Importance and challenges of prompt engineering
• The role of prompts in AI systems
• Principles of effective prompt design
• Prompt optimization techniques
• Applying prompts in NLP

Module 10: Studying Autoencoders and Generative Models

• How autoencoders work (structure and types: Vanilla, Denoising, Variational, Sparse, and Convolutional)
• Training and optimizing autoencoders
• Using autoencoders for dimensionality reduction, anomaly detection, and image restoration
• Generative AI fundamentals
• Key concepts and applications of GANs (Generative Adversarial Networks)
• Building a GAN model with TensorFlow 2.x
• Using GPT models
• Building a Q&A chatbot with Hugging Face

Module 11: Mastering Diffusion Models and Image Generation

• How diffusion models work (types: DDPM, Latent Diffusion Models, and Score-Based Generative Models)
• Training and optimizing diffusion models
• Applications (ultra-realistic image generation, super-resolution, video generation, and molecular biology)

Module 12: Leveraging LangChain and Large Language Models (LLMs)

• Introduction to LangChain and its ecosystem
• Key components of LangChain
• Text ingestion and chunking techniques
• Embeddings and vector databases
• LangChain Expression Language
• Building a simple RAG-based chatbot (Retrieval-Augmented Generation)

Module 13: Addressing AI Ethics

• General issues related to AI
• Specific ethical concerns in AI (bias, discrimination, etc.)
• The importance of ethics, bias mitigation, and building trust in AI

Module 14: Preparing for the AI Expert Exam (AI3050)

• Review of key course points
• Mock exam with corrections and explanations
• Exam strategies and tips for success (time management, reading techniques, and stress management)