SLIDE 1
Framework for Training BCI System
Ilya Kuzovkin
MTAT.03.277 Research Seminar in Data Mining University of Tartu, 2012
Outline
Background The Task Usual approach & difficulties Our attempt to improve things Results
Framework for Training BCI System Ilya Kuzovkin MTAT.03.277 - - PowerPoint PPT Presentation
Framework for Training BCI System Ilya Kuzovkin MTAT.03.277 - - PowerPoint PPT Presentation
Framework for Training BCI System Ilya Kuzovkin MTAT.03.277 Research Seminar in Data Mining University of Tartu, 2012 Outline Background The Task Usual approach & difficulties Our attempt to improve things Results
SLIDE 2
SLIDE 3
Background
Neurons transmit information via electrical signaling.
SLIDE 4
Background
Different areas of the scalp have different electric potentials We put electrodes and register these differences
SLIDE 5
The Task
Distinguish different mental states and map them to actions.
SLIDE 6
Usual approach & difficulties
Signal Processing
SLIDE 7
Usual approach & difficulties
Machine Learning
Instance (row) joint spectrums from all
channels at the moment of time
Attribute (column) certain frequency power
- n certain channel
Class specifies the instance label
SLIDE 8
Usual approach & difficulties
Signal is noisy
- Skull
- Skin
- Hair
- We can’t measure signal at it’s origins
Lack of data? No. Bad algorithms? No. Signal is not consistent for same mental state Test subject does not know how different his thoughts
are in terms of resulting electrical signal
SLIDE 9
Our attempt to improve things
We want to find a way to make learning process duplex:
not only machine will learn from user input, but user will learn and adapt his behavior according to feedback he receives from a machine
This can be formulated as establishing a
communication protocol between two agents during the interaction
- General Idea
Goal is to find set of messages which can be clearly formulated by human and distinguished well by machine
SLIDE 10
Our attempt to improve things
One Particular Realization The Signal Clustered signal instances Cluster distribution Resulting cluster distribution Interaction
SLIDE 11
Our attempt to improve things
Implementation Details Reading signal in real time & Signal Processing
Same as was explained earlier
Exploring cluster distribution Exploration vs. Exploitation
- Only explore – distribution will
randomly change every time
- Only exploit – there is a good
chance we missed best mental state and without exploration we will never find it Relaxed Imagine flower Fear or frustration
SLIDE 12
Our attempt to improve things
Current solution: K-means, e-greedy strategy Results Initial Distribution
This is how it looks like after one minute of thinking about random things
Relaxed
Test subject was trying to be as relaxed as possible, not concentrating on anything particular
Anxious
Test subject was performing calculations in the mind
SLIDE 13
Conclusion
Done
- Skeleton solution which can be used to test strategies
- Implementation of basic strategy
- Learning a bit about nature of distributions in different mental states
To Do
- More careful signal pre-processing
- Learn more about cluster consistency (longer trials, averaged results, different
test subjects)
- Find clustering algorithm which will fit nicely under multi-armed bandit
formalization of exploitation-exploration trade-off (other than k-means)
- Find best exploration-exploitation trade-off strategy for our type of data
SLIDE 14