DASC5335 - Deep Learning

DASC 5335 -- Deep Learning

Updated January 16, 2024

Office and Addresses

Delta 171 Phone 281.283.3805
email: boetticher@uhcl.edu
Secretary: Ms. Caroline Johnson, Delta 161 281.283.3860

Class Hours (Face-to-Face or Online)

Wednesday 7:00 - 9:50 PM, Delta Building: 237, or via Zoom (If necessary)
Zoom information may be found in the Google folder for this course.

Office Hours

Wednesday 1 PM to 4 PM; Thursday 9 to 10 AM, or by appointment. Students with appointments have priority. If the suite door is locked, then call my extension (x3805) using the phone in the hallway. Students who have an appointment will have priority over those students who don't

Teaching Assistant

Mr. Naga Sai Venkatesh Perumalla
Email: perumallan1637@uhcl.edu

TA Hours: Monday 7 - 10 PM; Tuesday 10AM to 3 PM; 7 - 10 PM; Wednesday 7 - 10 PM

If you are not willing to learn, no one can help you.

If you are determined to learn, no one can stop you. -- Zig Ziglar

Course Description

Deep learning. Neural Networks (NN), Recurrent Neural Networks (RNN) and Convolutional Neural Network (CNN) and their applications in various domains. Theory, design, implementation and optimizations of different neural networks. Neural networks architecture. Build, train and apply fully connected deep neural networks. Practice in Tensorflow. Laboratory instruction.

The traditional graduate student load is 3 courses. Be prepared to commit 15 to 20 hours per week to this course!

Course Goals

By the end of the course, you will

Understand the data mining process.
Have a working knowledge of different data mining tools and techniques.
Have an understanding of various Machine Learners (ML).
Have a working knowledge of some of the more significant current research in the area of data mining and ML.
Be aware of various data mining data repositories for the study of data mining.
Be able to effectively apply a number of data mining algorithms (e.g., neural networks, genetic algorithms) to solve data mining problems from various problem domains including Financial and Bioinformatics.
Be familiar with several successful applications of data mining.

Prerequisites

A graduate course in Data Mining or Machine Learning. If you do not meet the prerequisites, then you need to drop this course!

Methodology

Lecture, seminar, case studies, and interactive problem solving.

Appraisal:

Midterm, Final, Term Project (2 highest grades count 32% each. Other counts 16%)	80%
Quizzes and Participation	5%
Homework	15%

Grades will be based solely on criteria listed above. No other factors will be considered.

Grading Scale

93+ = A; 90 = A-; 87+ = B+; 83+ = B; 80+ = B-;

77+ = C+; 73+ = C; 70 = C-; 67+ = D+; 63+ = D; 60+ = D-; 0+= F

My motto:

Foster disciplined, altruistic passion.

Required Textbook

Glassner, Andrew. Deep learning: a visual approach. No Starch Press, 2021.

Reference Books

Atienza, Rowel. Advanced Deep Learning with Keras: Apply deep learning techniques, autoencoders, GANs, variational autoencoders, deep reinforcement learning, policy gradients, and more. Packt Publishing Ltd, 2018.

Ballard, Will. Hands-on deep learning for images with TensorFlow: build intelligent computer vision applications using TensorFlow and Keras. Packt Publishing Ltd, 2018.

Beysolow II, Taweh . Introduction to Deep Learning Using R: A Step-by-Step Guide to Learning and Implementing Deep Learning Models Using R. Apress, 2017.

Brownlee, Jason. Deep learning with Python: develop deep learning models on Theano and TensorFlow using Keras. Machine Learning Mastery, 2016.

Chollet, Francois. Deep learning with Python. Simon and Schuster, 2021.

Godoy, Daniel, Deep Learning with PyTorch Step-by-Step: A Beginner's Guide: Volume I: Fundamentals

Goodfellow, Ian, Yoshua Bengio, and Aaron Courville. Deep learning. MIT press, 2016.

Michelucci, U. Applied Deep Learning—A Case-Based Approach to Understanding Deep Neural Networks; Apress Media, LLC: New York, NY, USA, 2018. ISBN 978-1-4842-3789-2.

Millstein, Frank. Convolutional neural networks in Python: beginner's guide to convolutional neural networks in Python. Frank Millstein, 2020.

Moolayil, Jojo, Jojo Moolayil, and Suresh John. Learn Keras for deep neural networks. Birmingham: Apress, 2019.

Nielsen, Michael A. Neural networks and deep learning. Vol. 25. San Francisco, CA, USA: Determination press, 2015.

Trask, Andrew W. Grokking deep learning. Simon and Schuster, 2019.

Zhang, Aston, et al. "Dive into deep learning." arXiv preprint arXiv:2106.11342 (2021). (Free!)

Deep Learning Software

Keras
PyTorch and TensorFlow
Perceptilabs
Weights & biases
openAI/gym

Schedule (Tentative)

Jan 17 - Introduction to deep learning, Applications of Deep Learning, Neural Networks Basics - 01

************************************************************************

*** All course materials are located in the Google Drive folder. ***

*** You are expected to bring a copy of the notes to all lectures. ***

*** I strongly recommend you place the notes in a 3-ring binder. ***

************************************************************************

FOR THIS WEEK (IF NOT SOONER)

Blue Color = Available on the Google Drive

· Read: Deep Learning, A visual approach - Chapter 13

· Read: WK01 Notes - Introduction to Deep Learning

It is the student's responsibility to download the notes, print the notes, and bring them to class.

· Read: Syllabus

FOR NEXT WEEK (IF NOT SOONER)

· Read: Deep Learning, A visual approach - Chapter 14 - Backpropagation

· Read: WK02 - Notes - Mathematical Foundation (Calculus and Tensors)

Jan 24 – Mathematical Review (Calculus, Tensors), Term Project discussion

Assignment 1 - Calculus Tensor Problems

Point value: 100 points

Due date: Wednesday, Feb 14 at 7^PM. Email to boetticher@uhcl.edu

FOR NEXT WEEK (IF NOT SOONER)

· Read: WK03 Notes - Activation Functions

Jan 31 – Neural Networks Basics - 02 (Activation Functions)

FOR NEXT WEEK (IF NOT SOONER)

· Read: WK04 Notes - Loss (Cost) Functions

· Read: WK04B - Automatic Differentiation

Feb 07 – Neural Networks Basics - 03, (Loss Functions)

Assignment 2 - FeedForward Neural Network Analysis

Point value: 100 points

Due date: Wednesday, February 28 at 7^PM. Email to boetticher@uhcl.edu

FOR NEXT WEEK (IF NOT SOONER)

· Read: Deep Learning, A visual approach - Chapter 15 - Optimization

· Read: WK05 Notes - Optimization

Feb 14 – Neural Networks Basics - 04 (Optimization Regularization, Dropout, Backpropagation)

Assignment 1 is due. Email to boetticher@uhcl.edu

Assignment 3 - Classical NN Optimization Analysis - Part 2

Point value: 100 points

Due date: Wednesday, March 6 at 7^PM. Email to boetticher@uhcl.edu

FOR NEXT WEEK (IF NOT SOONER)

· Read: WK06 Notes - Pulling it all together

Feb 21 – Pulling it all together, Regularization and Dropout

FOR NEXT WEEK (IF NOT SOONER)

· TBD

Feb 28 – Term Project Proposals, Midterm Review

Assignment 2 is due. Email to boetticher@uhcl.edu

FOR NEXT WEEK (IF NOT SOONER)

· Submit: Midterm questions by Tuesday, March 5^th, 7 PM.

Use the template found on the Google Drive

Strip out any identifying information (Your name, Student ID number)

Specify whether you want me to post your questions with your peers

· Study!

Mar 06 – Midterm (Closed Book, Closed notes)

FOR NEXT WEEK (IF NOT SOONER)

· Read: Chapter 16 Convolutional Neural Networks

· Read:

FOR NEXT WEEK (IF NOT SOONER)

· Read: Deep_Learning A Visual Approach - Chapter 16 - CNN

· Read: Deep_Learning A Visual Approach - Chapter 17 - ConVets in Practice

Mar 13 ************ Spring Break *************

Mar 20 – Convolutional Neural Networks (CNNs) Architectures (AlexNet, VGG)

FOR NEXT WEEK (IF NOT SOONER)

· Read: Deep Learning A Visual Approach - Chapter 22 - GANs

· WK09 Notes - GANs

Mar 27 – Generative Adversarial Networks (GANs)

FOR NEXT WEEK (IF NOT SOONER)

· Read: Deep Learning A Visual Approach - Chapter 21 - Reinforcement Learning

· WK10 Notes - Reinforcement Learning, Graph Neural Networks, Federated Learning

Apr 03 – Reinforcement Learning. Graph Neural Networks, Federated Learning

FOR NEXT WEEK (IF NOT SOONER)

· Read: Deep Learning A Visual Approach - Chapters 19, 20 - RNNs, Transform Networks

· WK11 Notes - Recurrent Nerual Networks, LSTM

******** April 09 – Last day to withdraw ********

Apr 10 – Recurrent Neural Networks (RNNs), Modern Recurrent Neural Networks (GRU, LSTM,... )

FOR NEXT WEEK (IF NOT SOONER)

· Read the Week 12 notes and papers

Apr 17 - Capsule Networks, Bidirectional Encoder Representations from Transformers, Chat GPT3

FOR NEXT WEEK (IF NOT SOONER)

Apr 24 – Term Project Presentations

FOR NEXT WEEK (IF NOT SOONER)

· Submit: Final questions by Tuesday, May 2^nd, 7 PM. This is optional.

Use the template found on the Google Drive

Strip out any identifying information (Your name, Student ID number)

Specify whether you want me to post your questions on the Google Drive.

· Study!

May 01 – Final Exam

Other Policies

This class has 6 simple rules:

1) Be respectful of others. (3% Penalty/Cellphone or text infraction)

2) Be very passionate about your learning and do your best.

3) Be fearless - ask lots of questions in class.

4) Don't be late on anything. (10% Penalty/Day)

5) Don't ever cheat.

6) Have fun!

Miscellaneous

Any person with a disability who requires a special accommodation should inform me and contact the Disability services office or call 281 283 2627 as soon as possible.
You are expected to come fully prepared to every class!
If there is any religious observance that may interfere with any scheduled exam, homework due date, or attending class, please notify me of the situation during the first 2 weeks of class so that adjustments can be made at that time.
Please turn off all cell phones, and pagers prior to the start of class.
I am willing to provide letters of recommendation/references only if you have attained an 'A' in one of my classes, or two 'A-' in two of my classes.
I highly recommend that you seek out your advisor and complete your Candidate Plan of Study (CPS) as soon as possible. I am normally not available for advising during the summer months.

Return to Top

Home • UHCL • SCE

2700 Bay Area Boulevard
Delta Building. Office 171
Houston, Texas 77058
Voice: 281-283-3805
Fax: 281-283-3869
boetticher@uhcl.edu