Course List and Descriptions
Brief course description: This course covers the following topics: Search: uninformed search, informed search and multi-agent search. Machine learning techniques, including linear regression, logistic regression, decision trees, random forests, Bayesian networks, inference and sampling, hidden Markov models and particle filtering; K-nearest neighbors, K-means, support vector machines, gradient boosting. Neural networks and optimization. Introduction to computer vision and natural language processing. Reinforcement learning and recommendation systems. Generative models.
Brief course description: This course covers the principles of machine learning, data preparation and loading, basic vector operations, multi-layer perception, convolutional neural networks, sequential prediction, vision recognition, text recognition, model optimization and deployment, and distributed training. The students will learn these principles through a number of design and implementation coursework to enrich their system development skills.
Brief course description: This course introduces emerging and enabling technologies in artificial intelligence to solve real-world problems in a various daily and industrial application. The course will offer a number of research or topical seminars, given by research or industry experts, to let the students keep abreast of the trends of the development in artificial intelligence technologies.
Brief course description: This course provides an introduction to deep learning. Students taking this course will learn the basics, theories, models, algorithms, implementation, and recent progress of deep learning, and obtain empirical experience on the designing, training, and understanding of deep neural networks.
Brief course description: This course will cover fundamental knowledge and advanced topics in image processing, computer vision and pattern recognition, including image acquisition, processing and transformations, feature detection, segmentation, motion estimation, stereo vision, 3D reconstruction and modeling, scene detection and classification, object alignment and tracking, face detection and recognition, color image processing and digital photography. Machine learning and deep learning algorithms for computer vision and their system applications will be discussed.
Brief course description: This course introduces the main concepts of the natural language processing tasks, such as parsers, syntax, semantics, discourse, language modeling, machine translation, text generation and summarization, multilinguality and multimodality, etc. It also covers the research methods, including statistical and stochastic learning methods, expectation maximization algorithms and deep learning models. It will further discuss the applications and technological advances in large language models.
Brief course description: This course covers mathematical and computational foundations of generative modeling, as well as applications in engineering, design, science, and the arts. The topics include foundation models, deep learning models, variational autoencoders, generative adversarial networks, autoregressive models such as Transformers, neural text decoding, prompt programming, and detection of generated content. Applications to language models and multimodality system, as well as social responsibility in generative modeling will also be discussed.
Brief course description: This course introduces the principles and techniques of Trustworthy Artificial Intelligence (Trustworthy AI), which aims to mitigate the potential adverse effects of AI on people and society. The course focuses on four main aspects of trustworthy AI: privacy & security, robustness, explainability, and fairness. It covers the state-of-the-art research progress in these areas, including federated learning and adversarial attacks. Algorithms, models, and systems will be covered. Moreover, the course discusses the ethical and social implications of trustworthy AI, to foster social awareness among students who would use or develop AI techniques in the future. This course is suitable for students who have some background in machine learning, probability, and linear algebra.
Brief course description: The high computational demands of deep neural networks (DNNs) coupled with their pervasiveness across both cloud and IoT platforms have led to a rise in specialized hardware and software techniques to accelerate DNN executions. This course will present techniques that enable efficient applications and computing of DNNs. The course will start with an overview of DNNs, and then will introduce various frameworks and architectures that support DNNs, as well as the implementations and optimizations on some particular computing platforms.
Brief course description: This course first introduces fundamentals of machine learning with a large size of samples, including basic principles (maximum likelihood vs least redundancy) and typical structures (linear systems of hidden factors, mixture of local structures, and Markov temporal models), The second part of the course covers learning theories towards small sample size challenge, including major topics (model selection, learning regularization, two stage implementation, sparse learning, and automatic model selection) and three streams of efforts, namely generalization error estimation (CV, AIC, VC theory), shortest coding length (MML vs MDL) or similarly various Bayes (BIC, MAP, Laplace, marginal, and variational), and BYY learning (BYY system, best harmony theory, Ying-Yang alternation updating, and five action circling implementation).
Brief course description: This course introduces a dozen of machine learning algorithms and typical applications in business intelligence, natural language processing, computer vision, and sensor-based data analyses, including four topics that consist of (1) supervised learning algorithms induced by structural risk minimization for classification and regression problems (decision trees, logistic regression, support vector machines, regularized linear regression, kernel machines, etc.), and their applications in sensor-based indoor localization, business intelligence; (2) supervised learning algorithms based on deep learning (CNN, RNN, etc.), and their applications to natural language processing and computer vision; (3) unsupervised learning algorithms for clustering and representation learning (K-means, spectral clustering, autoencoder, etc.); (4) introductions of other learning algorithms and applications, such as transfer learning, recommender systems, sensor-based activity recognition, etc.
Brief course description: The course is designed for students who already have a background in deep learning and builds upon the knowledge gained in related introductory courses. It begins with a comprehensive coverage of the Transformer model, which has emerged as a universal and flexible learning architecture. Students will learn how state-of-the-art foundation models are constructed for different application domains using the Transformer as the basic building block. The course also covers how to leverage distributed / parallel computing infrastructure and methodologies to enable the training, serving, and deployment of foundation models in a scalable manner. Key aspects of foundation models including their emergent behavior, scaling laws, in-context learning ability, adaptation and augmentation are covered. The role of foundation models in the low-code, rapid development and adoption of new AI applications, as well as their societal considerations will also be discussed.
Brief course description: The course begins with a detailed exposition of probabilistic graphical models, then proceeds with various inference methods, including variational inference, belief propagation, and Markov Chain Monte Carlo (MCMC). In the second part of the course, we then discuss the connections between probabilistic models and risk minimization, as well as how optimization-based methods can be used in large-scale model estimation. Finally, the course will briefly discuss nonparametric models, e.g., Gaussian processes, and their use in practical applications.
Brief course description: This course aims to cover the fundamental topics relevant to Reinforcement Learning (RL), a computational learning approach where an agent tries to maximize the total amount of reward it receives while interacting with the complex and uncertain environments. The course content includes the basics of Markov Decision Processes, model-based and model-free RL techniques, policy optimization, RL distributed system design, as well as the case studies of RL for game playing such as AlphaGo, traffic simulation, and other robotics applications.
This course provides a comprehensive introduction to the fundamental theories, methods, and applications of multimodal machine learning. It covers core concepts related to multimodal machine learning, including unimodal (such as images, text, video, audio, sensor data) and multimodal representation, alignment, reasoning, generation, and transference. These concepts include, but are not limited to, multimodal transformers, multimodal tensor fusion, cross-modal alignment, multimodal interaction with reinforcement learning, multimodal generation, neuro-symbolic models, mutual information, and multimodal graph networks. We will also review recent papers describing state-of-the-art probabilistic models and novel algorithms for Multimodal Machine Learning and discuss the current and upcoming challenges.
Brief course description: This course introduces the advanced machine learning techniques for extending the boundary of life sciences. Topics to be covered include but are not limited to the recent successful stories of AI for life sciences study. We are going to cover reinforcement learning-based drug design (GENTRL), graph neural network-assisted antibiotics discovery, deep learning-enhanced super-resolution microscopy, ground-breaking molecular folding algorithm (AlphaFold), deep learning-based disease diagnosis and prediction, deep learning for single-cell and spatial transcriptomics, multi-modality/omics learning, model interpretability, privacypreserving learning for life sciences. Along the course, we will further discuss the challenges and opportunities of AI in life sciences, such as foundation models with unlabeled data for solving data scarcity issue, out-ofdistribution learning for drug design. This is a research-orientated course. By the end of the course, the students are expected to finish a substantial project, aiming at publication.
Brief course description: This course introduces the advanced topics in perception for intelligent robotics. It covers fundamental concepts and techniques of machine vision, robotic image and video processing, sensor fusion for semantic mapping and exploration, pattern recognition, learning and deep neural networks, robotic scenario intelligence, perception and anticipation of human behaviors, and advanced robotic trajectory and task planning. Case studies of successful medical and service robotics are discussed. In the course project, students are required to propose, design and implement a robotic system with intelligent perception to map and explore dynamic environment and interact with human subjects.
Brief course description: Introduction to wearable technology, reviews on wearable robotics, wearable sensor principles, wearable augmentation and machine intelligence, wearable design by "MINDS" (Miniaturization, Intelligence, Networking, Digitization, and Standardization), wearable medical devices and systems, wearable electro-physiologies, implantable therapeutic systems, sensor informatics, data-driven intelligent applications, and project topics of current interests.
Brief course description: This course covers the most up-to-date research in the field of neuromorphic hardware. The course will begin with an introduction to neuromorphic computing, followed by discussions on neuromorphic hardware development, in which the students will study the materials, devices, and circuits exploited to enable neuromorphic hardware. In addition to lectures, the course includes lab sessions where students can have experience with neuromorphic hardware fabrication.
Brief course description: The content of this course is composed of five sections: 1) Fundamentals of AI and typical algorithms; 2) Technology and system architecture of AI IC, as well as design methodology considering algorithm, software and hardware; 3) Software tool chains related to the application of AI IC; 4) Emerging semiconductor devices, computing modality, circuit design, system architecture, requirement and trend; 5) Practice of AI IC design. The students to be enrolled in this course are required to be with essential knowledge in AI algorithm, software and IC design.
Brief course description: In this course, students will learn fundamental image processing techniques, characteristics of different types of medical images, and how to apply different classical image processing techniques to different types of medical images. Topics covered in this course include but are not limited to:
- An overview of medical imaging modalities and their clinical use,
- Introduction to medical image computing, including registration, segmentation, classification, reconstruction, super-resolution, and visualization,
- Traditional image processing techniques for medical image analysis,
- Machine learning/deep learning for medical image analysis, and
- Frontline of AI in medical imaging and case studies.
Traditional AI models can respond to prompts or execute predefined tasks impressively. Agentic AI goes beyond traditional AI by incorporating a "chaining" capability: it can take a “sequence” of actions in response to a single request, breaking down complex tasks into smaller, manageable steps or coordinating among multiple agents to achieve a common goal. Agentic AI systems have a high degree of autonomy and can act with minimal human supervision. Agentic AI could revolutionize business and financial technologies of the future. This course introduces the students to the fundamentals of agentic systems, covering reinforcement learning agents, language model agents, and multi-agent systems. We will cover applications of agentic systems in various aspects of business and fintech, including cryptocurrency/blockchain, financial analysis, portfolio management, customer service, human resources, business planning, and operations.
Brief course description: This is a systems course that will enable students to have in-depth understanding of key information processing algorithms and their implementation for Internet of Things (IoT) systems. The topics cover 1) overview of basic signal processing algorithms such as FFT and digital filters; 2) advanced information processing algorithms such as acoustic and visual signal processing, spatial sensing, machine learning etc.; 3) their implementation on cutting-edge IoT platforms and key system issues of such as energy efficiency and realtime in the contexts of a set of key IoT applications such as smart health, environmental monitoring, smart homes/buildings, smart cities etc.
Brief course description: This course aims to introduce various key concepts, technologies and applications in edge artificial intelligence (AI), which is the implementation of AI applications in an edge computing environment, which allows the computation to be processed close to where the data is located. The topics covered include, but are not limited to: basic concepts and fundamental technologies in edge AI, hardware for edge AI (e.g., embedded AI accelerators, FPGA, and infrastructures), interaction between edge AI and the cloud, machine learning techniques for model compression, security and privacy in edge AI, applications of edge AI for smart health and autonomous driving.
This course introduces key AI technologies and their applications and systems in fundamental scientific fields. It covers the basics of AI, deep learning models, training methods, and the latest technology frameworks. The course delves into the interdisciplinary applications of AI in areas such as earth sciences, physics, chemistry, biology, and neuroscience, presenting recent advancements, technical challenges, and future trends. Focusing on these five major application areas, the course topics include weather forecasting, molecular property prediction, protein structure prediction, and neuron reconstruction. Through this course, students will not only understand the fundamental concepts and principles of AI in scientific research but also develop the ability to integrate AI with scientific practice.
Brief course description: This course will cover core ideas and advanced topics of computational imaging systems and algorithms, including camera and image sensor models, high dynamic range imaging, coded imaging systems (aperture, exposure, illumination), burst photography for low-light imaging, 3D imaging, plenoptic functions and light field, Neural Radiance Fields (NeRF), compressive sensing, neuromorphic imaging, optical neural network, and more. Emphasis is on novel hardware and system design of computational cameras, as well as solving inverse problems with classic optimization algorithms and modern end-to-end learning-based methods. Students will learn the core principles of many computational imaging systems and implement key optimization-based and learning-based algorithms to solve inverse problems.
Brief course description: The main goal of this course is to comprehensively introduce the principles and technical methods used in modern robot learning and control techniques. This course will not only teach students how to use data-driven methods to design, control and optimize robot systems through theoretical learning and practical operations, but will also delve into relevant basic knowledge and skills, to ensure that students can fully master and apply these important skills. Practical assignments will give students the opportunity to implement these techniques on simulated and real robot systems, providing them with a comprehensive educational experience.
This course will include the following main topics: robot kinematics and dynamics, optimal control, reinforcement learning in robots, teaching learning, embodied intelligent learning, trajectory optimization and robot perception, human-robot interactive and collaborative learning used in robot control and perception, machine learning, etc.
Brief course description: Human-computer interaction using speech and language processing draws from the perspectives of speech science, linguistics, modeling techniques and algorithms from engineering, as well as system implementations. This course presents fundamentals in speech production and perception, phonetics and phonology and signal processing techniques for speech analysis. We will then proceed to discuss the probabilistic approaches and statistical methods for speech and language processing, with particular emphasis on automatic speech recognition and text-to-speech synthesis technologies, as well as integrated spoken language systems. The course will also cover aspects of pattern recognition that are impactful to the practice and implementation of information processing systems.
Brief course description: Data mining provides useful tools for the analysis, understanding and extraction of useful information from huge databases. These techniques are used in business, finance, medicine and engineering. This course will introduce the techniques used in data mining. Topics will include clustering, classification, estimation, forecasting, statistical analysis and visualization tools.
Brief course description: This course aims at teaching students the state-of-the-art big data analytics, including techniques, software, applications, and perspectives with massive data. The class will cover, but not be limited to, the following topics: advanced techniques in distributed file systems such as Google File System, Hadoop Distributed File System, CloudStore, and map-reduce technology; similarity search techniques for big data such as minhash, locality-sensitive hashing; specialized processing and algorithms for data streams; big data search and query technology; managing advertising and recommendation systems for Web applications. The applications may involve business applications such as online marketing, computational advertising, locationbased services, social networks, recommender systems, healthcare services, or other scientific applications.
Brief course description: The aim of this course is to equip students with essential data analytics techniques for problems in financial technology. Topics covered by this course include linear and general regression, classification techniques such as logistic regression and discriminant analysis, decision trees, support vector machines, principal component analysis, clustering methods, time series models, resampling methods and deep learning models. Various applications in asset management, risk management, asset pricing and financial prediction will be used to illustrate the methods throughout this course. Students also learn how to implement these methods in the popular data analytics software R to analyze financial data.
Brief course description: This course gives an overview of social media, studies how different tools in information science can be used to analyse social media content, and how these results can be useful in different applications. This course will cover both knowledge in computational analysis of social media and the corresponding social networks, and basic technical skills in Python programming for the analysis.
Brief course description: This course aims to provide students an understanding in the operating principles and hands-on experience with mainstream Big Data Computing systems. Open-source platforms for Big Data processing and analytics would be discussed. Topics to be covered include:
- Programming models and design patterns for mainstream Big Data computational frameworks;
- System Architecture and Resource Management for Data-center-scale Computing;
- System Architecture and Programming Interface of Distributed Big Data stores;
- High-level Big Data Query languages and their processing systems.
Brief course description: This course aims at challenging the students to unleash their innovative power to create new AI applications and services that may lead to the formation of new startups in the Great Bay Area. The course will introduce a market-oriented model of thinking to complement the technology-oriented model, thus enabling the students to develop an all-rounded thinking pattern and get prepared to join the commercial world.
Brief course description: This course offers a unique blend of theory and practical application in transforming business ideas in artificial intelligence (AI) innovations into investable ventures, through holistic experiential learning experience. It empowers students with the strategies and communication skills needed to create compelling investment proposals, to convey their vision, articulate the unique value proposition of their AI solutions, and demonstrate their understanding of market dynamics and competitive landscape.
Brief course description: This course is designed to allow students to acquire a basic understanding and the skills of the practical aspects of the artificial intelligence area. During the internship, the student must attach to a company in a study-related position for no less than 12 weeks. The student will have an academic supervisor (as primary supervisor) and an industry co-supervisor from a company to provide advice to the student. The student must submit a report summarizing the internship experience at the end of the internship.
Brief course description: The students will design, research and develop projects under the supervision of an academic staff in related area of artificial intelligence. At the end of the course, the students must give presentations to the academic and/or industry supervisors, and submit project reports detailing the findings and deliverables.
Brief course description: The students will continue and further develop the project designs achieved in AIM5790 to address more functional and practical issues and make the projects more comprehensive and sustainable. The projects will be supervised by an academic staff in related area of artificial intelligence. At the end of the course, the students must give presentations to the academic and/or industry supervisors, and submit project reports detailing the findings and deliverables.