Santander Scholarship

  • Summary
    This research scholarship (£1,200 over 6 weeks) offers an enthusiastic 2nd year student a unique opportunity to work with a cross-country, cross-university, team of researchers from both academia and industry. The scholarship is funded by Santander under the Global Challenges Research Scheme.

    In the past year, COVID-19 detection via smartphone recorded coughs has emerged as a low-cost and non-invasive solution, which plays a vital role in supporting test-and-trace, and accelerating the return to pre-lockdown freedoms. However, there are challenges in cough classification via smartphones, including arbitrary-length audio segments, varying coughing intensity, and the high amount of ambient noises.

    This project will implement a well-calibrated Machine Learning algorithm to address the uncertainty of the critical COVID-19 patient detection task using smartphone recorded coughs.
    • image
    • image
    • image
  • Benefits
    Research publication: Providing the participating student with a strong piece of research record on their career portfolio. This placement is part of a larger research project funded by Brighton’s Radical Futures, Connected Futures initiative, and Santander. Therefore, the results will likely lead to a high quality publication in the premier Machine Learning and healthcare conference or journal.

    Research experience: Empowering the student with a set of transferable research skills, including paper reading, data analysis, machine learning algorithm implementation, parameter fine-tuning, and manuscript writing.

    Career opportunity: Offering the student an exclusive opportunity to collaborate with an industry practitioner, Cardisio (Germany), specialised in machine learning for healthcare, which will boost the student’s future employment with other corporations in the machine learning and healthcare industry.
  • Objectives & Requirements
    Through-out this intensive 6 week project, the student will be given clear, comprehensive tasks, while allowing some extra space and time for their own creativity, as follows.
    • Exploratory analysis on two COVID-19 audio datasets.
    • Build an algorithm to automatically segment audio clips into individual coughs.
    • Extract audio features and identify impactful ones.
    • Implement Conformal Prediction to classify the cough clips.
    • Analyse the effect of isolating cough samples for COVID-19 classification.
    Student's requirements.
    • Good Python programming knowledge.
    • Interest in Machine Learning.
    • Willing to accept constructive criticism.
  • Personnels
    Project leads: Dr. Khuong An Nguyen & Ms. Julia Meister (University of Brighton, UK).

    Project consultants: Prof. Zhiyuan Luo (Royal Holloway University of London, UK) & Mr. Werner Gentzke (Cardisio, Germany).
  • Dates & Applying process
    Application deadline: 31-May-2021.

    Project starting: 1-July-2021.

    Project ending: 12-August-2021.
    Step 1: Please send the latest CV and a covering letter explaining the reason for applying, by email to Dr. Khuong An Nguyen (K.A.Nguyen@brighton.ac.uk).

    Step 2: Register on Santander website: https://app.becas-santander.com/en/program/urs-global-challenges-student-applications
    "COVID-19 patient detection via smartphone recorded coughs and confidence Machine Learning".

    Step 3: An interview will be organised as soon as thereafter.