Course Webpage: CSCI 5922
Course Instructor: Prof. Michael C Mozer
This repository contains my solutions to assignments given in this course. Here is a brief summary of assignments with their respective links:
- Assignment 1: To implement linear regression and linear-threshold classification using scikit-learn. Dataset used: -random data-
- Assignment 2: To implement a one-hidden-layer back propagation network to process real data without using Tensorflow / any other library. Dataset used: Room occupancy detection data set.
- Assignment 3: Exploring tensorflow. Implement Hinton Family Trees. Dataset used: Family Relations
- Assignment 4: To use tensorflow to explore convolutional nets and to optimize a network for CIFAR-10 dataset.
- Assignment 5: The goal of this assignment is to look through the literature, find interesting research, and write a commentary about one paper that you found that you feel is noteworthy. I chose the paper The Neural Conversational Model.
- Assignment 6: Implementing vanilla RNN and LSTM to learn a parity operator. Executed on Summit Haswell Supercomputer at CU Boulder's Research Computing. Used SLURM workload manager. Code credits: Tensorflow-Examples.
- Assignment 7 Final Project: Deep Knowledge Tracing on Fractions with Shirly Berends, Ph.D. student at CU Boulder; Sean Kelly, VP Engineering and Co-Founder at Woot Math; Brent Milne, VP Research and Co-Founder at Woot Math Simbulus, Inc.. Thanks to Prof. Mike Mozer and Mohammad Khajah @mmkhajah. Based on the work by: Code by M. Khajah, Code by Chris Piech, Paper by M. Khajah, R. Lindsey and M. Mozer at CU Boulder and Paper by C. Piech, J. Bassen, J. Huang, S. Ganguli, M. Sahami, L. Guibas, J.S.Dickstein from Khan Academy, Google and Stanford University.
Feel free to email me for comments/questions -> [email protected]