9.641J | Spring 2005 | Graduate

Introduction to Neural Networks

Assignments

Late Submission Policy

Any homework submitted after class on the due date will be subject to a 20-point deduction per 24 hour period. Extensions should be requested at least 3 days in advance and will only be granted for exceptional reasons (e.g., conference submission).

ASSIGNMENTS SUPPORTING FILES
Problem Set 1: Integrate-and-Fire Neurons (PDF)  
Problem Set 2: Perceptrons and Linear Networks (PDF)  
Problem Set 3: Nonlinear Networks (PDF)  
Problem Set 4: Lyapunov Functions (PDF)  
Problem Set 5: Nonlinear Network Theory Again (PDF)  
Problem Set 6: Deconvolution and Antisymmetric Networks (PDF) telescope.mat (MAT)
Problem Set 7: Models of Associative Memory (PDF)  
Problem Set 8: PCA, k-means, and Ring Network (PDF) faces.mat (MAT)

kmeans.mat (MAT)

Problem Set 9: Delta Rule and Gradient Descent (PDF) ps9faces.mat (MAT)

perceptron_delta.m (M)

mnistabridged.mat (MAT)

Problem Set 10: Backpropagation, and Trajectory Learning (PDF) ps9faces.mat (MAT)

mnistabridged.mat (MAT)

Course Info

As Taught In
Spring 2005
Level
Learning Resource Types
Lecture Notes
Problem Sets