San
Francisco Bay Area Chapter of American Statistical Association
Monthly Seminar
|
Time: |
4:30 – 6:00 pm, Sep 20, 2018 (Thursday) (4:30-5:00 networking and light refreshment,
5:00-6:00 seminar) |
|
Location: |
Blakeslee Room, 9th floor of Thornton Hall, San
Francisco State University 1600 Holloway Ave, San Francisco,
94132 (see parking info below) |
|
Speaker: |
Dr. Alexandra Piryatinska Associate Professor Department of Mathematics, San
Francisco State University Email: alpiryat@sfsu.edu |
|
Title: |
Binary classification of multichannel-EEG
records based on the ε-complexity of continuous vector functions |
Abstract:
A methodology for binary
classification of mental states using EEG records is proposed.This
methodology is based on the theory of the ε-complexity of continuous
functions which is extended here to the case of vector-functions. This
extension permits us to handle multichannel EEG records. The
essence of the methodology is to use the ε--complexity coefficients
as features for classification of different types of vector-functions
representing EEG-records corresponding to different mental states. The
methodology contains two steps. The first step is an estimation of
the ε-complexity coefficients of the original signal and its finite
differences. The second step is utilizing Random Forest (RF) or Support Vector
Machine (SVM) classifiers. We demonstrated the performance of our method
on simulated data. We also applied it to the problem of classification of
multichannel-EEG records related to a group of healthy adolescents (39
subjects) and a group of adolescents with schizophrenia (45 subjects). We
found that the random forest classifier provides a superior result. In
particular, out-of-bag accuracy in the case of RF was 85.3%. Using 10-fold
cross-validation (CV), RF gave an average accuracy of 84.5% on a test set,
whereas SVM gave an accuracy of 81.07%.
We note that the highest
accuracy on CV was 89.3%. To compare our method with the classical approach, we
performed classification using the spectral features. In this case, the best
performance was achieved using seven-dimensional feature space, with an average
accuracy of 83.6%. The obtained results indicate the effectiveness of the
proposed methodology.
Parking
SF State provides parking for students, faculty, staff and
visitors. The cost to park at different facilities and parking locations for
each type of user is located at https://parking.sfsu.edu/sfsu-parking/campus-parking. See the parking map for the locations of
each of these lots available at https://parking.sfsu.edu/sites/default/files/maps/614_color_map_PARKING_0.pdf. A valid daily or semester permit is required
at ALL times in every lot. After 5:00 pm and on weekends, all lots are open to
the public.
The gymnasium parking structure, Lot 6, is the closest to the
meeting registration area. On weekends the structure is open to the public,
however the lot may fill quickly. Parking fees are US$5 for
two hours or US$8 for a day of parking. Daily parking permits
expire at midnight.
Lot 20, located near Maloney Field, is the largest general paid
public visitor parking lot. The parking garage is accessible on State Drive
from Lake Merced Boulevard. The parking lot is open 24-hours a days, 7 days a
week. Fees include US$5 for two hours, US$8 for
the day (daily permits expire at midnight) and US$16 for two
days. Level 4i Orange to 4n Orange and the roof level of Lot 20 are restricted
to staff and faculty between 7:00 am and 5:00 pm Monday through Friday.
Lot 25,
located off of Winston Drive, is also a general paid public visitor parking
lot. The parking lot is open 7:00 am to 10:00 pm, 7 days a week. Fees
include US$5 for two hours, US$8 for the day
(daily permits expire at midnight) and US$16 for two days.
Please note that exact change is required;
machines do not give change. Pay station will accept $1, $5, and $10 bills as
well as credit/debit card. If credit/debit card is used there is a 2.75% credit
card fee.
Stonestown Shopping Mall parking lots are only for patrons
shopping within the mall. Stonestown has posted their parking policies
throughout the lots and has a towing company monitoring the areas. Violators
will be towed away at vehicle owner’s expense, per CVC 22658.A.
For additional parking information, please contact the parking
office at (415) 338-1441 or email parking@sfsu.edu.