Generative AI: Fundamentals to Advanced Techniques

• Introduction to Machine Learning

• Different Learning Approaches

• Unsupervised Learning: Correlation Analysis

• Unsupervised Learning: Clustering, K-Means Algorithm

• Supervised Learning: Regression, Linear Regression

• Supervised Learning: Classification, KNN Algorithm

• Introduction to Artificial Neural Networks

• Supervised Learning: ANN as a Classifier

• ANN's Advantages and Disadvantages

• Introduction to Deep Learning

• Review of ANNs

• Deep Convolutional Neural Networks

• DCNN: Regression with Tabular Data

• DCNN: Classification with Tabular Data

• DCNN: Classification with Images

• Other Deep Models

• Introduction to Generative AI and Transformers

• Introduction to Transformer Architecture

• Neural Networks, Encoder-Decoder Structure

• Introduction and Types of Attention Mechanisms

• Mathematical Foundations of Transformers

• Transformer Components: Embedding Layers

• Transformer Components: Self-Attention Mechanism

• Transformer Components: Multi-Head Attention

• Transformer Components: Feedforward Networks

• Transformer Training: Data Preparation and Training Strategies

• Applications: NLP, CV and Others

• Transformer Model Extensions and Hybrid Models

• Recent Innovations and Future Trends

• Introduction to Transformer-Based Models: BERT and GPT

• BERT: Model Architecture

• BERT: Pre-training and Fine-tuning

• BERT: Applications and Case Studies

• GPT: Model Architecture

• GPT: Pre-training and Fine-tuning

• GPT: Applications and Case Studies

• BERT and GPT: Implementation and Best Practices

• Model Combinations and Extensions for Specialisation

• Performance Optimisation

• Basics of Reinforcement Learning: Markov Decision Process, Value Function, Value-State Function, Policy

• Exploration vs Exploitation Dilemma

• Bellman Equations

• Dynamic Programming

• Monte Carlo Learning and Temporal Difference Learning

• SARSA and Expected SARSA

• Q-Learning and Function Approximation

• Policy Gradient (PPO)

• Reinforcement Learning for Human Feedback (RLHF)

• RLHF Implementation and Case Study; ChatGPT

• Direct Preference Optimisation (DPO)

• Gen AI Model Mechanics

• Autoencoders

• Variational Autoencoders (VAEs)

• GANs

• Training VAEs

• Training GANs

• VAEs and GANs for Image Generation

• Introduction to Diffusion Processes

• Diffusion Model Architecture

• Training Diffusion Models

• Sampling from Diffusion Models

• Performance Optimisation for Diffusion Models

• Applications of Diffusion Models

• Introduction to Computer Vision Models

• Transformer Architecture in Computer Vision

• Vision Transformer (ViT)

• Contrastive Language-Image Pre-Training (CLIP)

• Fundamentals of Multi-Modal AI

• Text-to-Image Generation Models

• DALL-E and Stable Diffusion

• API Integration

• Prompt Engineering for Image Generation

• Fine-Tuning for Specialised Tasks

• Model Combination

• Real World Applications and Case Studies

• Ethical Considerations

• Model Limitations

• Fundamentals of LangChain

• Using LangChain for Conversational AI

• Implementing Chains in LangChain

• Introduction to Retrieval Augmented Generation (RAG)

• Integrating RAG for Performance Optimisation

• Working with Documents and Vector Embeddings

• Introduction to Multi-Agent Systems

• Types of Multi-Agent Systems

• Communication Protocols

• Coordination Strategies

• Multi-Agent Reinforcement Learning

• LLMs as Agents

• Task Planning and Execution

• Problem-Solving Applications

• Challenges and Future Trends

The programme is designed for professionals with a minimum diploma and experience in object-oriented programming. It's ideal for those looking to gain expertise in AI, Machine Learning (ML), and deep learning, particularly in Generative AI, to advance their careers in tech, AI, and ML programmes, ML projects, or AI-driven industries.

The curriculum spans Python programming, machine learning programme content (supervised/unsupervised/reinforcement learning), neural networks, transformer models (BERT, GPT), diffusion models, multi-modal AI, and ethical considerations. It blends deep learning and artificial intelligence with hands-on AI ML certification projects and practical deep learning programme elements.

This 14-week (8-10 hours/week ) programme offers a focused and practical deep learning AI programme. It’s perfect for working professionals looking to upskill through a high-impact deep learning certification.

The programme is delivered entirely online, featuring both live sessions from the NUS School of Computing faculty and industry experts. Pre-recorded content supports flexible learning, while live interaction enhances understanding, a hallmark of top-tier AI certificate programmes and AI certification online experiences.

Participants will use tools such as TensorFlow, PyTorch, and LangChain while completing a capstone project focused on solving real-world challenges using AI with deep learning. These machine learning projects are designed to simulate scenarios from AI, Machine Learning (ML), and deep learning AI programme applications in industries including finance and healthcare.

You can expect to invest 8–10 hours per week over 14 weeks. This includes video lectures, assignments, and live discussions, mirroring the structure of a machine learning online programme or an artificial intelligence online programme from a top global university.

Yes. You can pay the full fee upfront or choose between two or three instalments. This flexibility supports learners globally. The installment plans and financial support options are available. Participants can contact the programme advisors for more information.

Yes. Participants will collaborate with global peers during live sessions and group projects. This not only builds connections but also deepens understanding of AI, ML, and deep learning concepts, offering a community like that of a structured AI certification programme or AI programmes for beginners.

A minimum diploma and familiarity with object-oriented programming (e.g., Python) are required. Prior experience in tech, data science, or AI and ML programmes is beneficial but not mandatory for joining this AI certification path.

You’ll have access to the learning platform and resources for 12 months after programme completion, valuable for review and keeping pace with the rapidly evolving field of AI and deep learning.

Taught by world-class faculty from the NUS School of Computing, this programme combines theoretical foundations with real-world practice, including tools like Stable Diffusion and GPT-4. It’s one of the best AI programmes available, merging ethical AI practices, technical mastery, and project-based learning, akin to AI certification programmes offered by elite institutions.

We encourage learners to complete the course to fully understand the concepts and derive valuable learning outcomes. However, if you still choose to withdraw, you may request a full refund within 7 days of payment or 14 days after course commencement, whichever comes later. After this period, the course fee becomes non-refundable.