Assignments of EEE482 - Computational Neuroscience

Topics

Assignment 1: Linear Algebra and Probability Theory

• Solving linear systems of equations
• Pseudo-inverse calculation
• Sparsest and Least Norm solutions to a linear system
• Calculation of common statistics
• Confidence Interval and Standard Error calculation
• Resampling methods: Bootstrapping and Jackknifing
• Reverse inference via Bayesian Statistics
• Multivariate Normal Distribution
• Standard Deviational Ellipse

Assignment 2: Neural Models

• Neuroelectronics
• Leaky Integrate-and-Fire Neuron Model
• Spike Triggerred Average (STA)
• Lateral Geniculate Nuclei cells and their Neural Activity
• Analysis of Neural Response
• Difference of Gaussians center-surround receptive fields
• Gabor receptive fields
• Image processing and Convolution

Assignment 3: Supervised Learning

• Ordinary Least Squares
• Linear & Higher Order Linearized Models
• Parametric & Nonparametric Nonlinear Models
• Maximum Likelihood Estimator
• Ridge Regression
• Cross Validation
• Hypothesis Testing

Assignment 4: Supervised/Unsupervised Learning

• Principal Component Analysis (PCA)
• Independent Component Analysis (ICA)
• Nonnegative Matrix Factorization (NNMF)
• Classical Multi-dimensional Scaling (CMDS)
• k-means Clustering
• Winner Take All Decoder
• Maximum Likelihood Decoder
• Maximum-a-Posteriori Decoder
• Linear Discriminant Analysis (LDA) Based Classifer

Project: Visual Object Recognition

• Task: Building a decoder that can map fMRI brain activity to the actual visual stimulus category.
• Note: Data used in the projects is very large so it is not provided here, however the data can be downloaded from:
  -> must be named data
  -> must be named haxby
• Teammates: Hygerta Imeri, Utku Gorkem Erturk