x - SNN Nijmegen - Radboud Universiteit

signtruculentΒιοτεχνολογία

2 Οκτ 2013 (πριν από 3 χρόνια και 2 μήνες)

48 εμφανίσεις

SNN Machine learning

Bert
Kappen
,

Luc
Janss
,
Wim

Wiegerinck
,
Vicenc

Gomez,

Alberto
Llera
, Mohammad
Azar
,

Bart van den
Broek
, 2 vacancies

Ender
Akay
, Willem Burgers

Activities


Approximate inference


Graphical models


Analytical methods


Sampling method


Control theory


Approximate inference


Reinforcement learning


Interaction modeling


Neuroscience


Adaptive BCI


ECoG


Neural networks


Bioinformatics


Genetic linkage analysis


Genome
-
wide association studies


Missing person identification


Smart Research


Wine portal


Petro
-
physical expert system


Credit card fraud detection


Promedas


www.promedas.nl/live


UMCU


Promedu




Approximate inference


Control theory


Neural networks


ABCI


ECoG


GWAS


Promedas


Graphical models

What are probabilities given evidence:


Intractable for large number of variables: 2
n

for binary variables

Junction tree method

Complexity reduction: 2
n

!

2
k

(n=8, k=3)

State of the art for intermediate size problems

No solution for large problems

Approximate inference

Optimal control theory

Minimization of cost function:

error cost at
the end

cost to reach the
target

Optimal solution hard to compute

Optimal control theory

Optimal control as a sum over trajectories,
Kappen

PRL 2005

Linear Bellman equation:


efficient computation of optimal controls


linear superposition of solutions


qualitative different results for high and low noise

Efficient computation

Theodorou
,
Schaal
, USC 2009

linear superposition of solutions

Da

Silva,
Popovic
, MIT 2009

Q
ualitative different results for high and low
noise


Delayed Choice


Optimal control predicts when
to act


More noise means more delay

Results

small noise

large noise

Control
signal

Trajectory

Modelling neural networks with activity
dependent synapses


Dynamic synapses


Recurrent connectivity and
Dynamic
synapses

Associative memory

dynamical memories

Storage
capacity


Sensitivity to external stimuli.

Relation to up
-
down states and
powerlaws


Discussion

Marro
, Torres,
Mejias

a) Electrophysiological preparation in
pyramidal neurons (layer 5) for a
pairing experiment.


b) Pairing: several current pulses
(during 200 ms) in the pre and post
-
synaptic neuron (4
-
8 action
potentials, AP) are injected 30 times
each 20 s.


c) Before: the response to stimuli is
variable and noisy.


d) After: optimal response to the first
current pulse and there is a decrease
of response to the next pulses.


e) The effect of “pairing” is robust
and Hebbian.

Markram

and
Tsodyks
, nature 1996
: Dynamic synapses

Sanchez
-
vives
, McCormick 2000

(
a) Intracellular recording in the primary visual cortex of a
halothane
-
anesthetized cat reveals a rhythmic sequence of
depolarized and hyperpolarized membrane potentials. (
b
)
Expansion of three of the depolarizing sequences for
clarity. (
c
)
Autocorrelogram

of the intracellular recording
reveals a marked periodicity of about one cycle per three
seconds. (
d
) Simultaneous intracellular and extracellular
recordings of the slow oscillation in ferret visual cortical
slices maintained
in vitro. Note the marked synchrony
between the two recordings. The intracellular recording is
from a layer 5 intrinsically bursting neuron. The trigger
level for the window discriminator of the extracellular
multi
-
unit recording is indicated. (
e
) The depolarized state
at three different membrane potentials. (
f
)
Autocorrelogram

of the intracellular recording in (
d
) shows
a marked periodicity of approximately once per 4 seconds.

Phenomenological model: Tsodyks y Markram (1997)

Attractor neural networks

Associative memory with “static synapses” (
D
j
=
1
, F
j
=
1)


Hopfield network

Oscillations occur for P>1 and more realistic neuron models


(Pantic et al, Neural Comp. 2002)

Phase portrait

Storage capacity

Sensitivity to external stimuli: h
i
+
d
i
m

Stimulus grows

Discussion


Synapses
show a high variability

with a
diverse origin
: the stochastic opening of the
vesicles, variations in the Glutamate concentration through synapses or the spatial
heterogeneity of the synaptic response in the dendrite tree (Franks et al. 2003)
.


Due to synapse dynamics, the
neural activity

loses stability
which
increases the
sensitivity to changing external stimuli: the concept of dynamical
memories


Synaptic depression reduces memory capacity


Synaptic facilitation
improves short time memories


Adaptive BCI


A BCI device is called adaptive if it is able to
change during performance in order to
improve it.


Proposed approach: Use Error Related
Potentials to provide the device with
feedback about its own performance.

Llera
, Gomez, van
Gerven
, Jensen

General idea


Use error related
potentials to provide
the device with
feedback about its
own performance.


Are Error Related Potentials possible to
classify at the single trial level?



Experimental design


An MEG experiment have been designed to get insight
into error related fields in a BCI context.


The protocol has been carefully chosen to avoid
lateralization due to movement in the screen.


The protocol is intended to provide us with data
containing error related fields and minimal extra input.


Classification methods with best
results till now...


Transformation to

28 frequencies
in range 3
-
30 Hz.


Normalization
.


273 channels


6 time steps of 100 ms


150 trials per subject


Linear Support Vector
Machine


Illustration on toy data


One dimensional feature space.


Two Gaussian distributions, one for each class.


Learning boundary using delta rule each time that we detect an error potential
.


Since Error Potentials classification is not 100%, we can have two undesirable effects:


Not learn when we should (prob2).


Learn when we should not (prob1).


Assume that probability of errors is the same for both classes.


ECoG

connectivity patterns during a motor
response task


We

have

computed

the

brain

connectivity

patterns

associated

to

a

simple

motor

response

task

from

ECoG

data

recordings
:




Functional

connectivity

:

Gaussian

graphical

models










Effective

connectivity

:

Direct

Transfer

Function

(DFT),

Granger

causality
.





Frequency domain.

Provides a non
-
symmetric causal matrix.

Does capture time evolution.

Assumes a good fit of the MVAR model.

Time domain.

Provides a symmetric independence matrix.

Does not capture time evolution.

Assumes normally distributed residuals.

Gomez, Ramsey

ECoG

connectivity patterns during a motor
response task





104

electrodes

(
101

after

preprocessing)
.



Implanted on the left hemisphere.



Two days, 40 trials per condition per day.



Sampling Rate 512 Hz: 1792 samples per trial.



22 bits,
bandpass

filter 0.15


134.4 Hz).



Inter
-
electrode distance : 1 cm.




ECoG

connectivity patterns during a motor
response task


The

Gaussian

model

reveals

clusters

of

correlated

activity

and

significant

differences

between

stimulus

and

response

states

related

with

motor

areas
.

ECoG

connectivity patterns during a motor
response task




With Granger causality we are able to identify a set of source electrodes (red
dots) which drive another subset of target electrodes.




Sources are similar in both conditions, although targets differ for stimulus and
response conditions.

Bayesian

Variable

Selection
:
causal

modeling

or

prediction
?


Stochastic

search
multiple
-
regression

building (
using

Gibbs

sampling
algorithm

based

on

George &
McCulloch

1995).


Efficient

in
large

p

problems

(500K
predictors
)


Extended

in a
hierarchical

model to
estimate

shrinkage

parameter
from

the data,
which

we have
shown

to
avoid

overfit
.


Model
averaging

using

half
-
certain

associations

was
shown

to
improve

prediction

substantially
:

Janss
,
Franke
,
Buitelaar

Some

extensions

/ research topics


Use

of prior
information

to help
(
constrain
)
finding

interactions

between

predictors


Multi
-
phenotype

modelling

and
prediction

using

an

embedded

Eigenvector
decomposition

in a
Bayesian

hierarchical

model.


Multi
-
layered

variable

selection

to model
genetic

effects

on

brain

fMRI
,
which

models
cognitive

tasks

and
psychiatric

disorders

x
1

x
2

x
3

x
4

x
5

x
6

x
7

x
8

x
9


x
500000


y

y
1

y
2

y
3

y
4

y
5


Pathway

1

Pathway

2

x
1

x
2

x
3

x
4

x
5

x
6

x
7

x
8

x
9


x
500000


u
1

u
2

u
3


Interactions

selected

within

pathways

EV latent

vectors

x
1

x
2

x
3

x
4

x
5

x
6

x
7

x
8

x
9


x
500000


y
1

y
2

y
3

fMRI

data

per
voxel

Janss
,
Franke
,
Buitelaar

PROMEDAS


PRObabilistic


Medical


Diagnostic


Advisory System

Waarom?


Toenemende complexiteit van diagnostiek


Falen in medisch handelen


98000 patiënten in de VS sterven per jaar als gevolg van falend
medisch handelen


Foute diagnose is frequent (8
-

40 %)


Toenemende kosten gezondheidszorg



Beschikbaarheid van
elektronische

data




Input:


patiëntgegevens,
klachten, symptomen,
labgegevens


Output:


Diagnoses, suggesties
voor vervolgonderzoek


Gebruikers:


Huisartsen


Opleiding


Management


Medisch specialisten


Grafische modellen

Grafische modellen

Exponentiele complexiteit

10

1 sec

20

20.000 sec

30

15 year

40

300.000 year

50

10
10

year

Bomen zijn netwerken zonder lussen

De
berekening

is
zeer

snel

voor

bomen

Promedas

graaf

benaderen

als

boom

Message passing


Exact op bomen


Goed op netwerken met weinig lussen


Wordt slechter met


aantal lussen


verbindingssterkte


Medische inhoud


Interne
geneeskunde

voor

specialisten


4
000
diagnoses,

4000
symptomen
,

60000
relaties


informele

klinische

evaluatie


50 test patients


score of correct diagnoses in top 3


6 % all in the top 3


26 % two in the top 3


54 % one in the top 3


14 % not correct


Sinds

oktober

2008
geintegreerd

in UMCU.
Ongeveer

1200
sessies

per
maand
.


Toekomst plannen


Promedas wordt gecommercialiseerd door een nieuw bedrijf Promedas
BV.



Mogelijke markten:


Web applicatie of cd
-
rom


Geïntegreerd in een ziekenhuis informatiesysteem


Telemedicine


….



Projectteam


Algoritmes & software


SNN, Radboud Universiteit Nijmegen




Medische inhoud


UMC Utrecht


www.promedas.nl


Onderwijs

N & S
voor

deze

richting



Bachelor


Neural networks and
information theory


Neurofysica




Master


Machine Learning


Computational
Neuroscience


SNN Colloquia