XClose

Cortexlab

Home
Menu

Neuroinformatics class page

Materials for Kenneth Harris' yearly course on Neuroinformatics.

Lectures will be delivered  in person 11am-1pm Wednesdays in the ground floor lecture room of the Sainsbury Wellcome Centre and will be recorded for those accessing remotely - zoom links are on the class moodle page. Workshops will be at 1:30pm Wednesdays in the same place.

On Weds March 13 the lecture will be in lecture theatre G03, 26 Bedford Way.  

The final assessment will be Thursday 21st March, 2pm-5pm, in the SWC ground floor lecture room where the lectures took place.

 PowerpointsWorksheets (non-assessed)Matlab code*

Week 1
10/01/2024

Review of Statistics
Spike Trains
The t-test and randomization tests
Making a raster plot

CorrelationOfCorrelations
Overfitting
PlaceFields
Rasters

Week 2
17/01/2024

LFPs
Spectral analysis
Power spectra

PowerSpectrum
GaussianProcesses
Tapering

Week 3
24/01/2024

Spectrograms and non-stationary signals
Multiple Timeseries
Spectrograms and coherencePowerSpectrumParameters
Spectrograms
HilbertTransform
MeanPhase
 

Week 4
31/01/2024

Circular Statistics, local likelihood
The bootstrap
Circular statistics

BootStrapCircular

Week 5
07/02/2024

Singular Value Decomposition
Dimensionality Reduction
Singular value decomposition

SvdMovie
SVDspikes

Week 6Reading week  

Week 7
21/02/2024

Multiple linear regression
Ridge Regression and Bayesian regression

Multiple linear regression
RidgeRegression

Week 8
28/02/2024

Lasso, Support Vector Machines, GLMs
The kernel trick
Poisson regression 

Week 9
06/03/2024

Analysis of multiple spike trains
Nonsense correlations
  

Week 10
13/03/2024

Cluster analysis and spike sorting
Bayesian methods
  

Week 11
20/03/2024

Information theory
Estimating mutual information
Information theoryMutualInformation

*The Matlab code recreates some of the figures shown in the slides. Please note the extensions of these files have been changed from .m to .txt to comply with our web service. Once you download them, you need to change them back to .m to run them in Matlab.