This the multi-page printable view of this section. Click here to print.

Return to the regular view of this page.

Assignments

Assignments

Assignments

Due dates are on Canvas. Click on the links to checkout the assignment pages.

1 - Assignment 1

Assignment 1

Assignment 1

In the first assignment you will be writing a technical document on the current technology trends that you’re pursuing and the trends that you would like to follow. In addition to this include some information about your background in programming and some projects that you have done. There is no strict format for this one, but we expect 2 page written document. Please submit a PDF.

Go to Canvas

2 - Assignment 2

Assignment 2

Assignment 2

In the second assignment, you will be working on Week 1 (see @sec:534-week1) lecture videos. Objectives are as follows.

  1. Summarize what you have understood. (2 page)
  2. Select a subtopic that you are interested in and research on the current trends (1 page)
  3. Suggest ideas that could improve the existing work (imaginations and possibilities) (1 page)

For this assignment we expect a 4 page document. You can use a single column format for this document. Make sure you write exactly 4 pages. For your research section make sure you add citations to the sections that you are going to refer. If you have issues in how to do citations you can reach a TA to learn how to do that. We will try to include some chapters on how to do this in our handbook. Submissions are in pdf format only.

Go to Canvas

3 - Assignment 3

Assignment 3

Assignment 3

In the third assignment, you will be working on (see @sec:534-week3) lecture videos. Objectives are as follows.

  1. Summarize what you have understood. (2 page)
  2. Select a subtopic that you are interested in and research on the current trends (1 page)
  3. Suggest ideas that could improve the existing work (imaginations and possibilities) (1 page)

For this assignment we expect a 4 page document. You can use a single column format for this document. Make sure you write exactly 4 pages. For your research section make sure you add citations to the sections that you are going to refer. If you have issues in how to do citations you can reach a TA to learn how to do that. We will try to include some chapters on how to do this in our handbook. Submissions are in pdf format only.

Go to Canvas

4 - Assignment 4

Assignment 4

Assignment 4

In the fourth assignment, you will be working on (see @sec:534-week5) lecture videos. Objectives are as follows.

  1. Summarize what you have understood. (1 page)
  2. Select a subtopic that you are interested in and research on the current trends (0.5 page)
  3. Suggest ideas that could improve the existing work (imaginations and possibilities) (0.5 page)
  4. Summarize a specific video segment in the video lectures. To do this you need to follow these guidelines. Mention the video lecture name and section identification number. And also specify which range of minutes you have focused on the specific video lecture (2 pages).

For this assignment we expect a 4 page document. You can use a single column format for this document. Make sure you write exactly 4 pages. For your research section make sure you add citations to the sections that you are going to refer. If you have issues in how to do citations you can reach a TA to learn how to do that. We will try to include some chapters on how to do this in our handbook. Submissions are in pdf format only.

Go to Canvas

5 - Assignment 5

Assignment 5

Assignment 5

In the fifth assignment, you will be working on (see @sec:534-intro-to-dnn) lecture videos. Objectives are as follows.

Run the given sample code and try to answer the questions under the exercise tag.

Follow the Exercises labelled from MNIST_V1.0.0 - MNIST_V1.6.0

For this assignment all you have to do is just answer all the questions. You can use a single column format for this document. Submissions are in pdf format only.

Go to Canvas

6 - Assignment 6

Assignment 6

Assignment 6

In the sixth assignment, you will be working on (see @sec:534-week7) lecture videos. Objectives are as follows.

  1. Summarize what you have understood. (1 page)
  2. Select a subtopic that you are interested in and research on the current trends (0.5 page)
  3. Suggest ideas that could improve the existing work (imaginations and possibilities) (0.5 page)
  4. Summarize a specific video segment in the video lectures. To do this you need to follow these guidelines. Mention the video lecture name and section identification number. And also specify which range of minutes you have focused on the specific video lecture (2 pages).
  5. Pick a sport you like and show case how it can be used with Big Data in order to improve the game (1 page). Use techniques used in the lecture videos and mention which lecture video refers to this technique.

For this assignment we expect a 5-page document. You can use a single column format for this document. Make sure you write exactly 5pages. For your research section make sure you add citations to the sections that you are going to refer. If you have issues in how to do citations you can reach a TA to learn how to do that. We will try to include some chapters on how to do this in our handbook. Submissions are in pdf format only.

Go to Canvas

7 - Assignment 7

Assignment 7

Assignment 7

For a Complete Project

This project must contain the following details;

  1. The idea of the project,

Doesn’t need to be a novel idea. But a novel idea will carry more weight towards a very higher grade. If you’re trying to replicate an existing idea. Need to provide the original source you’re referring. If it is a github project, need to reference it and showcase what you have done to improve it or what changes you made in applying the same idea to solve a different problem.

a). For a deep learning project, if you are using an existing model, you need to explain how did you use the same model to solve the problem suggested by you. b). If you planned to improve the existing model, explain the suggested improvements. c). If you are just using an existing model and solving an existing problem, you need to do an extensive benchmark. This kind of project carries lesser marks than a project like a) or b)

  1. Benchmark

No need to use a very large dataset. You can use the Google Colab and train your network with a smaller dataset. Think of a smaller dataset like MNIST. UCI Machine Learning Repository is a very good place to find such a dataset. https://archive.ics.uci.edu/ml/index.php (Links to an external site.)

Get CPU, GPU, TPU Benchmarks. This can be something similar to what we did with our first deep learning tutorial.

  1. Final Report

The report must include diagrams or flowcharts describing the idea. Benchmark results in graphs, not in tables. Use IEEE Template to write the document. Latex or Word is your choice. But submit a PDF file only. Template: https://www.ieee.org/conferences/publishing/templates.html (Links to an external site.)

  1. Submission Include,

  2. IPython Notebook (must run the whole process, training, testing, benchmark, etc in Google Colab) Providing Colab link is acceptable.

  3. The report in PDF Format

This is the expected structure of your project.

In the first phase, you need to submit the project proposal by Nov 10th. This must include the idea of the project with approximate details that you try to include in the project. It doesn’t need to claim the final result, it is just a proposal. Add a flowchart or diagrams to explain your idea. Use a maximum of 2 pages to include your content. There is no extension for this submission. If you cannot make it by Nov 10th, you need to inform the professor and decide the way you plan to finish the class.

Anyone who fails to submit this by the deadline will fail to complete the course.

For a Term Paper

For a graduate student, by doing a term paper, the maximum possible grade is going to be an A-. This rule doesn’t apply to undergraduate students.

For a term paper, the minimum content of 8 pages and a maximum of 10 pages must include using any of the templates given in project report writing section. (https://www.ieee.org/conferences/publishing/templates.html (Links to an external site.))

So when you are writing the proposal, you need to select an area in deep learning applications, trends or innovations.

Once the area is sorted. Write a two-page proposal on what you will be including in the paper. This can be a rough estimation of what you will be writing.

When writing the paper,

You will be reading online blogs, papers, articles, etc, so you will be trying to understand concepts and write the paper. In this process make sure not to copy and paste from online sources. If we find such an activity, your paper will not be accepted. Do references properly and do paraphrasing when needed.

Keep these in mind, before you propose the idea that you want to write. The term paper must include a minimum of 15 references which includes articles, blogs or papers that you have read. You need to reference them in the write-up. So be cautious in deciding the idea for the proposal.

Submission date is Nov 10th and there will be no extensions for this. If you cannot make it by this date, you need to discuss with the professor to decide the way you want to finish the class. Reach us via office hours or class meetings to sort out any issues.

Special Note on Team Projects

Each member must submit the report. The common section must be Abstract, Introduction, Overall process, results, etc. Each contributor must write a section on his or her contribution to the project. This content must be the additional 50% of the report. For instance, if the paper size is 8 pages for an individual project, another 4 pages explaining each member’s contribution must be added (for the two-person project). If there are 4 members the additional pages must be 8 pages. 2 additional pages per author. If results and methods involve your contribution, clearly state it as a subsection, Author’s Contribution.

8 - Assignment 8

Assignment 8

Assignment 8

For term paper submission, please send us the pdf file to your paper in the submission.

If you’re doing a project, please make sure that the code is committed to the repository created at the beginning of the class. You can commit all before submission. But make sure you submit the report, (pdf) and the code for the project. Please follow the report guidelines provided under Assignment 7.

Please note, there are no extensions for final project submission. If there is any issue, please discuss this with Professor or TA ahead of time.

Special Note on Team Projects

Each member must submit the report. The common section must be Abstract, Introduction, Overall process, results, etc. Each contributor must write a section on his or her contribution to the project. This content must be the additional 50% of the report. For instance, if the paper size is 8 pages for an individual project, another 4 pages explaining each member’s contribution must be added (for the two-person project). If there are 4 members the additional pages must be 8 pages. 2 additional pages per author. If results and methods involve your contribution, clearly state it as a subsection, Author’s Contribution. Good luck !!!