Doctoral Course: Learning from Sensor Data

Wednesday, May 2, 2018 to Friday, May 4, 2018

Learning from Sensor Data

 

Venue:

Room: PR104

 

Schedule:

Wed 2 May: 9:00-12:00 and 13:00-16:00

Thu 3 May: 9:00-12:00 and 13:00-16:00

Fri 4 May: 9:00-12:00



Lecturer:

 
Behnaam Aazhang
Electrical and Computer Engineering
Rice University
 


Abstract 

 
This short course will offer graduate students depth in a few topics in the area of data-driven engineering. The course will focus on a probabilistic approach to working with sensor data. That includes characterizing statistical properties of data and designing a system based solely on data. Therefore, the framework to study the physical system will be data driven. Possible topics could include representation of data, estimating key characteristics of data, and powerful data processing tools. These tools will be used in inference, classification, and clustering problems. 

 

Syllabus

• A probabilistic approach
                                           • Statistical characteristics of data

                      • Frameworks for learning from data 
                                           • Parametric models
                                           • Non-parametric—data driven 
                      • Estimating key statistical metrics from data
                                           • Plugin estimators 
                                                                                       • Kernel density estimation (KDE)
                                                                                       • K nearest neighbor (k-NN)
                      • Data representation using graphical modeling
                                                                 • Directed graphs
                                                                                       • Bayesian network
                                                                 • Undirected graphs
                                                                                       • Markov random fields
                                                                 • Factor graphs

Passing the course

Participation in the lectures, homework problems and an exam

Add to calendar

Event location

PR104

Back to events

Last updated: 3.5.2018