| Departments |
Computer Science, Zienkiewicz Institute for Modelling, Data and AI
|
|---|
| Description | We are looking for highly motivated research students that would like to work on healthcare datasets. Nowadays, healthcare systems around the world experience difficulties trying to meet patient demands, with many patients being frustrated and ending up in poor health as a result. This is where AI and machine learning algorithms can have a tremendous potential for helping bringing the demand down.
This project proposes therefore to use statistical and AI techniques to discover new and subtle patterns that may be early signs of more or less serious conditions. These patterns may not be easily recognisable straightaway by experts in the area. The NHANES dataset will be the primarily source of data, but other datasets may be considered depending on funding availability. In the first part of this project, the research student will have to get use to the NHANES visualiser tool being currently developed at Swansea prior to adding their own innovative pattern-finding algorithms.
Code will be primarily implemented in Python, but prior PHP, MYSQL and Laravel knowledge would be useful.
Blood Sample Photo By: Kaboompics.com: https://www.pexels.com/photo/set-of-vials-and-test-tube-of-blood-4226912/
Data Science Photo by Lukas: https://www.pexels.com/photo/close-up-photo-of-survey-spreadsheet-590022/ |
|---|
| Preparation | Some background reading should be done in the areas of:
-Principal Component Analysis
-Deep neural networks (concepts)
-Autoencoders (concepts, network layout)
-Generative Adversarial Networks (GANS)
Possible Programming Languages and tools: Python and Tensorflow/Keras, C++, Javascript, CSS, html, php
Environment: Any web browser, Linux Windows
Suitable to Advanced Software Engineers
Reading and online resources :
-Machine learning (series of videos):
https://www.youtube.com/watch?v=mbyG85GZ0PI
https://alex.smola.org/drafts/thebook.pdf
-Autoencoders
https://towardsdatascience.com/applied-deep-learning-part-3-autoencoders-1c083af4d798
https://en.wikipedia.org/wiki/Autoencoder
https://www.jeremyjordan.me/autoencoders/
-PCA
https://www.cs.princeton.edu/picasso/mats/PCA-Tutorial-Intuition_jp.pdf (See Figure 2,3)
http://www.cs.otago.ac.nz/cosc453/student_tutorials/principal_components.pdf
Opendatasets:
nhanes
https://wwwn.cdc.gov/nchs/nhanes/
Exhaustive list of tests with some explanations:
https://www.rcpa.edu.au/Manuals/RCPA-Manual/Pathology-Tests
https://www.clinicallabs.com.au/functional-pathology-old/practitioners/guide-to-pathology-tests/
https://www.clinicallabs.com.au/media/2342/guide-to-path-tests-aclmar-bf-nat-00059.pdf
https://gps.camdenccg.nhs.uk/cdn/serve/service-downloads/1517312915-89ba5f5e4a0bd34817edcbb1d77a642c.pdf |
|---|
| Project Categories |
Artificial Intelligence (AI), Bioinformatics, Data Science |
|---|
| Project Keywords |
Health Informatics, Life Science, Statistical Analysis, Web Applications |
|---|
Level of Studies
|
| Level 6 (Undergraduate Year 3) |
yes
|
|---|
| Level 7 (Masters) |
yes
|
|---|
| Level 8 (PhD) |
yes
|
|---|
