-WE7506e93b81.png)
Embedded system and Deep Learning
AI & FinTech
This workshop covers two threads: Embedded system and Deep learning.
Embedded system thread will covers basic hardware / low level software interaction. You will learn how to interface with useful peripherals, e.g. sensors, actuators etc.
In the Deep Learning section, you will begin by looking at the fundamentals of statistical models like regression models, Bayesian classifiers, Decision Trees and Support Vector Machines. From there you will explore classical neural network learning algorithms gradient descent and unsupervised methods, before delving into contemporary deep learning methods like Convolutional Neural Networks, Recurrent Neural Networks, Long Short Term Memories, Generative Adversarial Networks and Autoencoders. You will take a hands-on approach and will learn to identify key features of the problem at hand, and choose appropriate deep learning architectures and strategies to solve those problems.
To bridge between the hardware and the deep learning back-end, you will learn about how to efficiently transfer data over Message Queueing Telemetry Transport (MQTT), RESTful APIs, and learn to store and manage that data using both SQL and NoSQL databases. Lastly you will learn to secure your communications by generating and signing cryptographic keys and certificates.
You will apply the ideas learned in an intensive 2-3 person teams to design and build a hardware + software system. The system is freeform and open ended, but should includes hardware interfacing and deep learning. For example, a "home security" system that use movement sensors and deep learning to understand the typical movement of the occupants. Any out of ordinary movements will trigger an alert.
Embedded system thread will covers basic hardware / low level software interaction. You will learn how to interface with useful peripherals, e.g. sensors, actuators etc.
In the Deep Learning section, you will begin by looking at the fundamentals of statistical models like regression models, Bayesian classifiers, Decision Trees and Support Vector Machines. From there you will explore classical neural network learning algorithms gradient descent and unsupervised methods, before delving into contemporary deep learning methods like Convolutional Neural Networks, Recurrent Neural Networks, Long Short Term Memories, Generative Adversarial Networks and Autoencoders. You will take a hands-on approach and will learn to identify key features of the problem at hand, and choose appropriate deep learning architectures and strategies to solve those problems.
To bridge between the hardware and the deep learning back-end, you will learn about how to efficiently transfer data over Message Queueing Telemetry Transport (MQTT), RESTful APIs, and learn to store and manage that data using both SQL and NoSQL databases. Lastly you will learn to secure your communications by generating and signing cryptographic keys and certificates.
You will apply the ideas learned in an intensive 2-3 person teams to design and build a hardware + software system. The system is freeform and open ended, but should includes hardware interfacing and deep learning. For example, a "home security" system that use movement sensors and deep learning to understand the typical movement of the occupants. Any out of ordinary movements will trigger an alert.
About Professor
Professor Colin Tan
Department of Computer Science, School of Computing, NUS
Prof Tan received his Ph.D. degree in Computer Science from the National University of Singapore. He has taught classes on embedded systems design, control system design, real-time operating systems, and mobile applications development. He has conducted research on unmanned aircraft for over 10 years in NUS.
His research is in autonomous control of Unmanned Aerial Vehicles, and has publications in prestigious conferences like the Guidance and Navigation Conference held by the American Institute of Aeronautics and Astronautics, and the International Conference on Autonomous Agents and Multiagent Systems (AAMAS).
Department of Computer Science, School of Computing, NUS
Prof Tan received his Ph.D. degree in Computer Science from the National University of Singapore. He has taught classes on embedded systems design, control system design, real-time operating systems, and mobile applications development. He has conducted research on unmanned aircraft for over 10 years in NUS.
His research is in autonomous control of Unmanned Aerial Vehicles, and has publications in prestigious conferences like the Guidance and Navigation Conference held by the American Institute of Aeronautics and Astronautics, and the International Conference on Autonomous Agents and Multiagent Systems (AAMAS).
Professor Soo Yuen Jien
Department of Computer Science, School of Computing, NUS
Prof Soo received the B.Sc., M.Sc. and Ph.D degree from NUS in year 2000, 2001 and 2006 respectively. He has since enjoyed teaching for 10 plus years to receive numerous teaching awards such as NUS Annual Teaching Excellence Award (2012/13, 2010/11, 2007/08), NUS Annual Teaching Excellence Award Honor Roll (2014); Faculty Teaching Excellence Award (2009/10, 2008/09, 2007/08, 2006/07); Faculty Teaching Excellence Award Honor Roll (2009/10). He has been inducted to the NUS Teaching Academy in 2012. His research interests are in computer organization and computer architecture.
Department of Computer Science, School of Computing, NUS
Prof Soo received the B.Sc., M.Sc. and Ph.D degree from NUS in year 2000, 2001 and 2006 respectively. He has since enjoyed teaching for 10 plus years to receive numerous teaching awards such as NUS Annual Teaching Excellence Award (2012/13, 2010/11, 2007/08), NUS Annual Teaching Excellence Award Honor Roll (2014); Faculty Teaching Excellence Award (2009/10, 2008/09, 2007/08, 2006/07); Faculty Teaching Excellence Award Honor Roll (2009/10). He has been inducted to the NUS Teaching Academy in 2012. His research interests are in computer organization and computer architecture.
