Research Assistant position in the area of reinforcement learning. In particular, the selected candidate will perform research activities related to the PEARL project (Provably Efficient and Adaptive Reinforcement Learning). The PEARL project aims to develop novel algorithms based on bisimulation for measuring the distance between two reinforcement learning problems. These algorithms can be used to construct compact representations of a problem by minimizing the bisimulation distance to an unknown, high-dimensional problem. This smaller problem can then be solved and the solution projected back to the large problem.
Principal camp de recerca / Main research field: Intel.ligència Artificial / Artificial Intelligence
Tasks to be performed:
Study of the state of the art in the field of reinforcement learning in general and bisimulation in particular. Concretely, study how bisimulation can be used to measure the distance between two reinforcement learning problems.
Development, implementation and evaluation of novel algorithms based on bisimulation for measuring the distance between reinforcement learning problems.
Study how these algorithms can be used to construct compact representations of a complex problem. Carry out experiments to compare the developed algorithm with the state-of-the-art. This activity is related to the workpackage WP2 of PEARL.
Contribution to the writing of deliverables and scientific publications that are developed within the framework of the PEARL project.
The duration of these tasks is estimated over a period of 12 months.
Main research field: Artificial Intelligence
Group and complement: Group 3 + Level s
Dedication and working hours: Part time (27,5 h/week)
Planned remuneration approx: 24K€ gross per year
Financing fundPRESP06324 - PGC (UE-PID2023-147145NB-I00) finançat per MCIN/AEI/10.13039/501100011033/ FEDER, UE - Agencia Estatal de Investigación (AEI), el Ministerio de Ciencia, Innovación y Universidades (MICIU) y el Fondo Europeo de Desarrollo Regional (FEDER) “Una manera de hacer Europa”
Requirements:
Batxillerat-FP2 / High school-FP2 or higher
The expected start date is 1 October 2026, the job is in Barcelona
Selection criteria: Candidate selection will be based on a CV evaluation and, if applicable, a test and/or interview. The assessment criteria are as follows:
1- Academic Training (0-40 points).
Degree and Master's degree in Engineering in areas related to artificial intelligence (mathematics, computer science etc.). In case the candidate has not completed the master's studies, it will be assessed whether he is doing the final master's thesis or similar, and its topic.
2- Other professional training and experience, adequacy to the proposed profile (0-40 points):
Extensive knowledge of artificial intelligence in general, and reinforcement learning in particular.
High level of programming in Matlab, Python and/or C.
Experience in using simulation/experimentation for research in artificial intelligence.
Proven experience in performance evaluation techniques, analysis, simulation and experimentation.
Extensive knowledge of mathematics.
3- Altres mèrits/ Other merits (0-20 points):
The minimum score to pass the selection process is 80 points. The candidate with the highest score in the selection process will be offered the job.
BASIC INFORMATION ON DATA PROTECTION:
Data Controller: Universitat Pompeu Fabra
Purpose: Research support staff recruitment.
Rights: You can access your data; request their rectification, deletion and portability; you may object to their processing and request their limitation.
Additional information: Further detailed information is available on the website: https://rat.upf.edu/ca/ll040
UPF promotes a diverse and inclusive environment and welcomes applicants regardless of age, disability, gender, nationality, race, religion or sexual orientation.
