Machine Learning for Verification

Machine Learning for Verification

Fully-Funded Doctoral Studentship in Formal Verification

Department of Computer Science, Oxford University

Supervisors: Professor Tom Melham and Professor Daniel Kroening 

Start Date: October 2018 or earlier.

Oxford University’s Automated Verification Research Group is offering a fully-funded D.Phil. (Ph.D.) studentship in the Department of Computer Science. 

The successful applicant will be working with a first-class team in Oxford’s world-leading verification research group, internationally recognized as one of the largest and strongest groups in the world.  Our work spans a wide range of research, from fundamental investigations into the decidability and complexity of model checking, through logics and semantic models, all the way to practical, machine-assisted methods applicable to real-world problems and programming languages. We also have exceptionally strong industrial links.

This position is associated with a research project funded by the Semiconductor Research Corporation (SRC) led by Professors Tom Melham and Daniel Kroening.  Through the project’s sponsorship by the SRC, the applicant will have the opportunity for engagement with top researchers and verification practitioners at SRC member companies; mentors for the project include leading verification scientists at Intel, ARM, and IBM.

This project is in the area of machine learning and formal verification for computer systems, such as devices in the Internet of Things.  The project aims to develop methods, algorithms, and tools for formal verification of systems by using machine learning to analyse component interactions based on data from execution traces.  The project will involve formal methods and theoretical foundations, and is aiming ultimately at practical results.  Potential applicants with a serious interest and suitable background are warmly welcomed to make informal enquires before applying, when further information about the project may be made available.

Within this broad scope, the exact directions pursued in the research will be flexible, to suit the background and interests of the candidate appointed.  Some areas of background experience that may have relevance to the project include the following:

• machine learning
• formal verification, theoretical or applied
• symbolic execution, model checking
• automated software verification
• analysis of concurrent systems
• internet of things applications

These topics are only indicative - we welcome all applicants with strong computer science background.

Candidates should also have good writing, communication, presentation, and organization skills.  Applicants must in addition satisfy the usual requirements for studying for a doctorate at Oxford. See the University’s web pages on the DPhil in Computer Science for details:

http://www.ox.ac.uk/admissions/graduate/courses/dphil-computer-science

 The studentship is funded at a level that will cover University and College fees at home/EU student rates, together with an annual stipend for three years of at least £14,553.  The studentships will be funded for three years and will start in October 2018, although an earlier start date may be possible. There will also be provision for some funds to support research travel during the doctorate. 

Applicants liable for fees at the overseas rate are welcome, but will have to make up the difference between home/EU and overseas fees.  Applicants who are citizens of qualifying countries (Taiwan, China as a whole, and Asia) may additionally be considered for the award of a Jason Hu Scholarship at Balliol College, which when combined with this studentship will cover all overseas University and College fees in addition to the additional stipend for living expenses.

Please apply  online  by 8 January 2018 quoting studentship reference code CS-TM-DK-2018.

If you have any questions about the studentship please email  Tom Melham .

If you have any questions about the studentship or the formal application University process please email   Julie Sheppard .