AMS 559: Smart Energy in the Information Age (Fall 2017)


Time & Location: TuTh 10:00AM - 11:20AM, HARRIMAN HLL 111


Instructor: Zhenhua Liu (PhD from Caltech CS, current research focuses on cloud computing, machine learning, optimization, algorithms, and energy)


Office hours: Tue/Thu 12-1pm @ Math Tower 1-118




This will be a research based course for graduate and senior undergraduate students.


This will be a research based course for graduate students and senior undergraduate students in special cases. The course has two sides: techniques and applications in energy. The techniques you learn will be applied to other problems.


On the technique side, we will go over popular techniques in machine learning (logistic regression, neural network, SVM, clustering, etc), optimizations, economics (mechanism design), and cloud computing.


On the application side, we will first introduce the issue in energy and sustainability, followed by recent applications of the techniques we learned on energy problems. In particular, machine learning is used for predictions of renewable generation, electricity prices, etc; optimization is used for optimal power flow problem and energy procurement; economics is used in demand response program design; and cloud computing is used as an important component of IoT.


If time allows, we will discuss some recent progress such as deep learning (AlphaGo, Libratus), as well as the applications of deep learning in energy. Examples include that Google leveraged deep learning to improve data center cooling efficiency, and DeepMind’s collaboration with UK National Grid.


Pre-requisites are reviewed in class to make the course as self-contained as possible. Anyone interested in this topic and/or my research is encouraged to enroll. As a research-based course, I need your resume/transcript to decide if your background fits and your SBU ID to have you enrolled. Meanwhile, don’t hesitate to send me an email if you have any questions.


Grading will be based on class presentation (20%), homework (30%), course projects (literature study and proposals, 40%), and class participation/interaction (10%). No exam or textbook required.


Tentative plan:





Week 1~2

Course introduction and overview

Week 3~6


Machine learning

Week 7~9

Optimal power flow


Week 10~11

Demand response


Week 12~13

Internet of Things

Cloud computing

Week 14

Data center cooling

Deep learning

Week 15

Final project presentation