EPSRC Winter School in Mathematics for Data Modelling

Factorial Switching Kalman Filters for Condition Monitoring in Neonatal Intensive Care

author: Chris Williams, University of Edinburgh

Description

This presentation describes the application of data modelling using Kalman filters to premature baby monitoring.

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Known Unknowns: Novelty Detection in Condition Monitoring
1:37	Premature Baby Monitoring
3:49	Why Model This Data?
4:19	Overview
6:08	Autoregressive (AR) Processes
8:24	Kalman Filter - 1
9:12	Kalman Filter - 2
10:24	Inference Problem: Filtering - 1
12:18	Inference Problem: Filtering - 2
13:58	Simple Example
16:03	Applications
16:46	Probes
17:31	Factors Affecting Measurements
19:10	Common Factor Examples - 1
19:55	Common Factor Examples - 2
20:33	- Questions
24:46	Factors Affecting Measurements
26:29	- Questions
30:39	- Questions
32:57	Factor Interactions - 1
33:23	Factor Interactions - 2
33:33	Factor Interactions - 3
33:48	Factor Interactions - 4
34:33	Factor Interactions - 5
36:06	Related Work
37:47	Kalman Filtering
37:51	Switching Dynamics
37:58	Kalman Filtering
38:05	Switching Dynamics
38:15	Factorial Switching Kalman Filter
38:43	Switching Dynamics
38:52	Inference
40:58	Gaussian Sum Approximation - 1
41:30	Gaussian Sum Approximation - 2
41:41	Gaussian Sum Approximation - 3
42:33	Parameter Estimation - 1
48:50	Parameter Estimation - 2
48:54	Parameter Estimation Example
49:58	Learning Stable Physiological Dynamics
50:25	- Questions
59:49	Inference Results - 1
60:27	Inference Results - 2
61:51	Quantitative Evaluation - 1
63:11	Quantitative Evaluation - 2
66:25	Quantitative Evaluation - 1
67:39	Comparison with FHMM Model
67:55	Novel Dynamics
69:11	Known Unknowns - 1
69:18	Known Unknowns - 2
69:29	X-Factor for Static 1-D Data
71:39	X-Factor with Known Factors
72:08	X-Factor for Dynamic Data
73:12	Spectral View of the X-Factor
74:05	X-Factor Demo
77:34	More Inference Results
77:46	EM for Novel Regimes
78:07	- Questions
86:06	- Questions
86:09	- Questions

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.