CE889-SP Neural Networks And Deep Learning Assignment

There is one assignment, by which you are expected to demonstrate the understanding of knowledge learnt from the module and the practice of machine learning models applied to a realistic classification problem.

1. Objectives

• Understanding the knowledge of neural networks and diverse models
• Being able to program to achieve models for a tailored context
• Being able to apply neural network models (including deep learning models) to a required application

2. Requirements

• Making the codes of the Logistic model for classifying the dataset (see the details of the dataset below). You must program by yourself using Python (show functions for each step of the classification).
• Selecting a deep learning model for the classification (Publicly available packages/toolboxes can be used [reference to the sources is required]. It is very encouraging to adapt the existing deep learning models for the purposes of performance improvement.)

3. Submission Files and Submission Deadline

Files: Report and CodesDate:
See the date shown on the submission system (Faser) or follow the announcement by School Office

4. Report Format

(No more than 10 pages for the whole report)

• Cover Page (Personal Information: Name, ID, Date etc.)
• Main Body
  • Abstract (a summary of the content)
  • Backgrounds (introduce relevant works, such as previous studies/papers/state-of-the-art relevant methods etc.)
  • Methods (a description of the methods)
  • Results (classification accuracies, accuracy comparison etc.)
  • Conclusions (a short summary of your results/findings)
• Reflections (no more than a half-page)
• References (papers, online webpages, books etc.)

5. Dataset

The dataset is available publicly and can be downloaded via https://bbci.de/competition/iv/#dataset2a

You are required to download Dataset 2a <4>. There are four classes (i.e., left hand, right hand, feet, and tongue). To download the dataset, you have to register with your email address. In the section of ‘Download of data sets’ on the webpage opened by the above link, you need to fill in the name and email address and will receive an email, in which the password and link for downloading are enclosed. Then, you can use your email address and password to log in and download data.

The detailed information can be found in the paperhttps://bbci.de/competition/iv/desc_2a.pdf

This dataset includes 9 subjects. In order to reduce the workload, you are only required to apply the Logistic Model and Deep Learning Model to the data of Subject 1 (data file named ‘A01T.gdf’).

6. Target

• Using 3-fold cross-validation to evaluate the classification performance of models. That is the trials contained in ‘A01T.gdf’ are equally divided into 3 sets. For each time, one set is for testing and the remaining two sets are for training. Each set will be the testing set once. After that, the average accuracy across the sets is the final classification accuracy.
• Obtaining classification accuracy using Logistic model
• Obtaining classification accuracy using a deep learning model (selected by yourself)
• Comparing the performance (accuracies) between models to understand the parameter tuning of deep learning models and how performance varies over the different settings.

Are you struggling to keep up with the demands of your academic journey? Don't worry, we've got your back! Exam Question Bank is your trusted partner in achieving academic excellence for all kind of technical and non-technical subjects.

Our comprehensive range of academic services is designed to cater to students at every level. Whether you're a high school student, a college undergraduate, or pursuing advanced studies, we have the expertise and resources to support you.

To connect with expert and ask your query click here Exam Question Bank