瑞士苏黎世大学2020年博士后职位招聘（统计学）

Job Description

UZH Digital Society Initiative

Post-Doctoral or Experienced Practitioner Position in Fairness of

Statistical Decisions 80-100 %

We are offering an interdisciplinary postdoctoral or experienced practitioner position on the topic Fairness of Statistical Decisions. The researcher will contribute to the project "Socially acceptable AI and fairness trade-offs in predictive analytics in recruitment and education/training". The position is advertised by the Institute of Biomedical Ethics and the History of Medicine in collaboration with the Digital Ethics Lab of the Digital Society Initiative, University of Zurich (UZH). Funding is provided for three years (extension possible) by the Swiss National Science Foundation within the National Research Program 77 "Digital Transformation".

Your responsibilities

The research projects involve a cutting-edge topic in digital ethics: fairness-by-design for AI (more broadly, for statistical decisions). The postdoctoral researcher is expected to work primarily with a moral/political philosopher (Dr. Michele Loi) to establish theoretical links between moral concepts like discrimination, moral justification and fairness/non- discrimination/de-biasing standards applicable in statistical modelling (ranging from the simplest statistical metrics, e.g. recall parity to the more conceptually sophisticated approaches, e.g. counterfactual fairness). The problem setting is defined by the so-called "impossibility of fairness," meaning the impossibility to satisfy multiple prima facie plausible statistical fairness desiderata at once. The question to be explored is "what kind of moral justifications can one offer, and codify, for or against different mathematical fairness desiderata, according to the context of choice?". The project is mainly one of translation: it seeks to establish how far ethical arguments formulated by a philosopher make sense to statisticians and how far ethical arguments formulated by a statistician make sense to a) a philosopher, b) other stakeholders, including human resource specialists and lay people. The team involves researchers from UZH, the Zurich University of Applied Sciences, the University of St. Gallen. You will be given leeway for developing own research goals within the framework of the research projects.

The projects are expected to have considerable practical implications, and are conducted in close collaboration with partners from the industry, in particular in the human resources domain. The researcher will be involved in the operational lead of a PhD student with a computer science background who is creating fairness-by-design and data visualization software and with an experimental philosopher / quantitative psychologists who works on fairness perceptions.

Your profile

We are open to considering two different profiles:

Profile 1: post-doctoral researcher (Ph.D. required)

PhD in philosophy of science (focus causality or probability) or cognitive science with a strong background in probability and statistics OR mathematics OR computer science OR statistical science OR similar.

Experience or at least ability to understand the fair ML literature (in particular, the probability/statistical aspects; ability to understand the computational aspects is desirable but non-mandatory).

A few publications in international journals, showing an inclination towards interdisciplinary and applied research, or at least potential for being developed in that direction.

Committed to write papers for a diverse range of audiences, including computer science, and either moral/political philosophy or philosophy of science.

Profile 2: experienced practitioner (no Ph.D. required)

Experienced researcher from the industry, practitioner of data science, in particular statistical modelling;

in a domain with social implications, where questions of discrimination emerge;

Knowledge of state-of-art fair ML approaches.

Both profiles:

Team spirit, commitment and pleasure in working with PostDocs and PhD students in the fields of computer science, philosophy, and management;

Strong interest in the social aspects of predictive modeling and autonomous systems, in particular fairness, non-discrimination and human rights;

Very good command of English, written and spoken;

Fluency in German would be an asset.What we offer

We offer competitive employment conditions at UZH, a highly engaging research environment at the Digital Society Initiative (DSI) and plenty of opportunities for interdisciplinary exchange. Salary will be according to UZH regulations.

The project is an excellent opportunity to further develop qualifications and experience in different disciplines at the interface of AI, autonomous systems and ethics.

Place of work

Zürich

Start of employment

Screening of applicants will start on March 1, 2020 and continue until the position is filled.

Please provide the following in a single pdf file when applying for the position:

Cover letter indicating the position you are applying to, stating your motivation, experience and skills you could bring to the project (2 pages max)

Detailed CV (no photo), including publications (3 pages max)

For profile 1: one page summary of your dissertation, two scientific articles (including your best article and the one most closely related to the topic of this job – be ready to explain their significance also to evaluators who may not belong to your academic field – does not need to be in a top journal, or even published)

For profile 2: a research statement (max 12.000 words) describing your motivations, the question you would like to explore, and the methods you envisage.

Contact information for two references.