European Molecular Biology Laboratory (Heidelberg) Featured PhD Programmes
University of Liverpool Featured PhD Programmes
King’s College London Featured PhD Programmes
Newcastle University Featured PhD Programmes
University of Sheffield Featured PhD Programmes

Understandability in Science

  • Full or part time
  • Application Deadline
    Thursday, April 04, 2019
  • Funded PhD Project (UK Students Only)
    Funded PhD Project (UK Students Only)

Project Description

One of the primary aims of physical scientists is to understand and explain nature. This project will investigate what this means in a world where the use of AI in science, and machine learning in particular, is now widespread; with mixed initiative discovery in which multiple participants (human and computer) work together towards a common goal, replacing our model of AI-as-tool with AI-as-collaborator.

The project will develop new approaches to understandability and explanation in science and mathematics. This may include the human-like computing movement in which machines are developed with human-like perceptual, reasoning and learning abilities in order to support collaboration and communication; the framing approach, in which software is enhanced with explanatory and argumentation functionality in order to mitigate the understandability problem; and the forgoing understandability approach, in which we decide to forgo understandability in science in return for the increases in power, generality and predictiveness that machine learning approaches give. The project will explore and further each of these approaches and develop new perspectives on understandability in science.

We anticipate that the main research question will lie in the framing approach: "Can automated discovery systems present and explain their results to an end user in ways that enhance understanding and perception of value?". Within the framework of explainable artificial intelligence (XAI), we will look at greater and lesser understandability in terms of describing processes underlying the generative act and consider these for ML approaches. This will be studied in the context of our application domain -- geology. In particular, geological rock interpretation provides a rich domain because (a) each rock represents a unique record, yet there are multiple interpretations that can be explained and argued for, and (b) it is possible to operate with an incomplete set of theories/concepts to generate a useful result.

Features of the project include:
- bringing together the fields of automated scientific discovery and argumentation
- working with geologists at every stage of design and development
- complementing a parallel PhD project in which ASD is applied to geology and enhanced with cognitive functionality (within the human-like computing approach)

Applicants wishing to apply should submit a one-page covering letter stating your background, academic qualifications (i.e. Masters Degree or Honours at 2:1 or above in a related subject), past research experience and interests, and future career aspirations. Please include a full CV, a copy of your academic transcript and the names and contact details of two referees to Dr Alison Pease, . Please also send any other informal inquiries or queries to the same email address.

Funding Notes

To be eligible for a fully-funded PhD studentship, covering tuition fees and an annual stipend set at UKRI rates, the candidate must have no restrictions on how long they can stay in the UK and have been ordinarily resident in the UK for at least 3 years prior to the start of the studentship (with some further constraint regarding residence for education, further guidance can be found on the EPSRC website). Due to funding requirements the University of Dundee is limited to accepting only UK students.

Related Subjects

Email Now

Insert previous message below for editing? 
You haven’t included a message. Providing a specific message means universities will take your enquiry more seriously and helps them provide the information you need.
Why not add a message here
* required field
Send a copy to me for my own records.

Your enquiry has been emailed successfully

FindAPhD. Copyright 2005-2019
All rights reserved.