[Forum SIS] PhD position OSLO - Explainable AI

[Arnoldo Frigessi]

PhD Position, 3 years, in statistics and machine learning.
University of Oslo, Department of Mathematics Oslo, Norway.

Theme of the PhD project: Explainable AI (XAI). 
XAI is a term used for the broad concept of making AI systems understandable to humans. A central topic is the ability to explain what causes a model to produce a certain outcome (prediction) in terms of a specific set of input features. High-end ML models often have a large number of input features. This complicates the explanation of the predictions, both computationally and in terms of presenting useful explanations. You will develop methodology for efficiently and accurately explaining predictions from any model when the features are high dimensional. The PhD research project will be performed in close contact with the Norwegian Computing Center and industrial and public partners in Big Insight, our center of excellence for research based innovation.

The PhD studies at the University of Oslo are characterized by a very strong research component: the PhD thesis usually includes several papers which are published in leading scientific journals, while the candidates are also required to take a small number of advanced courses as part of the training component of their degree.

Big Insight and the University of Oslo offer a lively and socially exciting environment, with generous funding for travelling to conferences and summer schools.

In Norway, PhD students are employed at the university giving the degree. Salary NOK 479600 – 523200 per annum depending on qualifications plus attractive welfare benefits and a generous pension agreement, in addition to Oslo’s family-friendly environment with its rich opportunities for culture and outdoor activities

Applicants must have completed their master studies by June 2020.

Deadline: 24th January 2020

More information is available here: 

https://www.jobbnorge.no/en/available-jobs/job/179885/phd-research-fellowship-in-statistics-and-machine-learning

Ingrid Glad