Job Reference
605_24_CASE_WP_RE1
Position
Research Engineer - ML-Aided Mathematical Optimization (MAMO) - (RE1)
Data de tancament
Dilluns, 30 Setembre, 2024
Reference: 605_24_CASE_WP_RE1
Job title: Research Engineer - ML-Aided Mathematical Optimization (MAMO) - (RE1)
About BSC
The Barcelona Supercomputing Center - Centro Nacional de Supercomputación (BSC-CNS) is the leading supercomputing center in Spain. It houses MareNostrum, one of the most powerful supercomputers in Europe, was a founding and hosting member of the former European HPC infrastructure PRACE (Partnership for Advanced Computing in Europe), and is now hosting entity for EuroHPC JU, the Joint Undertaking that leads large-scale investments and HPC provision in Europe. The mission of BSC is to research, develop and manage information technologies in order to facilitate scientific progress. BSC combines HPC service provision and R&D into both computer and computational science (life, earth and engineering sciences) under one roof, and currently has over 1000 staff from 60 countries.
Look at the BSC experience:
BSC-CNS YouTube Channel
Let's stay connected with BSC Folks!
We are particularly interested for this role in the strengths and lived experiences of women and underrepresented groups to help us avoid perpetuating biases and oversights in science and IT research. In instances of equal merit, the incorporation of the under-represented sex will be favoured.
We promote Equity, Diversity and Inclusion, fostering an environment where each and every one of us is appreciated for who we are, regardless of our differences.
If you consider that you do not meet all the requirements, we encourage you to continue applying for the job offer. We value diversity of experiences and skills, and you could bring unique perspectives to our team.
Look at the BSC experience:
BSC-CNS YouTube Channel
Let's stay connected with BSC Folks!
We are particularly interested for this role in the strengths and lived experiences of women and underrepresented groups to help us avoid perpetuating biases and oversights in science and IT research. In instances of equal merit, the incorporation of the under-represented sex will be favoured.
We promote Equity, Diversity and Inclusion, fostering an environment where each and every one of us is appreciated for who we are, regardless of our differences.
If you consider that you do not meet all the requirements, we encourage you to continue applying for the job offer. We value diversity of experiences and skills, and you could bring unique perspectives to our team.
Context And Mission
Mathematical optimization, aka integer programming (IP or ILP) problems are convex problems solved by discrete optimization, where the objective is the optimal allocation of given resources (represented by decision variables). The term "integer" is used to reflect the fact that resources are available for allocation in amounts subject to strict divisibility constraints.
In this topic a pure binary integer combinatorial optimization problem (ILP) is used to solve a fundamental question in Natural Language Processing (NLP): how to best compare and match the semantic contents of two natural language texts or "textual summaries" A and B, based on a given measure of the semantic similarity between their unit semantic constituents (USCs) ?
The student intern will focus on how to reduce the computational cost incurred when resolving pure binary integer linear programming (ILP) problems. The approach will consist in porting existing Python code to a high performance computing node partition, possibly using the PyCOMPSs framework for orchestration in distributed compute environments, and reducing the set of feasible ILP solutions by means of reinforced learning.
In this topic a pure binary integer combinatorial optimization problem (ILP) is used to solve a fundamental question in Natural Language Processing (NLP): how to best compare and match the semantic contents of two natural language texts or "textual summaries" A and B, based on a given measure of the semantic similarity between their unit semantic constituents (USCs) ?
The student intern will focus on how to reduce the computational cost incurred when resolving pure binary integer linear programming (ILP) problems. The approach will consist in porting existing Python code to a high performance computing node partition, possibly using the PyCOMPSs framework for orchestration in distributed compute environments, and reducing the set of feasible ILP solutions by means of reinforced learning.
Key Duties
- Become familiar with several concepts: NLP, the measure of semantic similarity, the extraction of unit semantic constituent from unstructured text, the ILP formulation. The student will also establish a baseline workflow with Python, based on a conventional ILP solver applied to our joint set-packing and setpartitioning problem, using one or more FOSS ILP Python packages, and following the already formulated constraints of the ILP problem.
- Explore how to parallelize the execution of the ILP algorithm using the task-based programming model COMP Superscalar ported to PyThon (PyCOMPSs) which eases the development of applications for the state-of-theart distributed infrastructure available at BSC (the 3rd most powerful and greenest in Europe, the 8th most powerful and 6th greenest in the world).
- Design a semantically informed, state-of-the-art Machine Learning method to reduce the set of feasible solutions for the initial semantic pairing ILP problem, whereby a reduced feasible set (or reduced feasible solutions space) is defined
as any possible ensemble of the most promising pairing combinations between semantic elements of text A and of text B, that satisfy the problem constraints. - Learn a generalizable model to infer acceptably reduced sets of feasible solutions for never-seen-before textual summaries, establishing suitable
performance metrics. - Relax the pure binary integer constraint on decision variables (variables that can only accept values of either 0 or 1), to a ternary integer constraint (variables that can accept values of either 0 or 1, or 2), and to a pure integer constraint (variables can accept any integer values within the range [0, nB - nA]) and will inspect obtained semantic similarity solutions from a natural (human) language
standpoint. - Prepare a situation and final report to be delivered on the last day of this short term contract. The report should include (i) a structured bibliography on salient questions relevant to the technical work to carry out, (ii) a description of methodologies used throughout the interneship, (iii) any developped code and results, and (iv) a prospective but well articulated report section on how you might envisage setting up an RL algorithm to learn a dimension-independent policy with the goal of reducing the feasible solution space of the ILP problem. You may include several possible ideas and describe them independently.
- Code developped will be stored in the https://gitlab.bsc.es/wavephenomenagroup/MAMO/ version-control project repository .
Requirements
-
Education
- Course subjects concentration in mathematics and computer science.
-
Essential Knowledge and Professional Experience
- Relevant scientific project experience gained through work experience or recent academic courses will be valued in the following areas: computer science, mathematical analysis and optimization schemes.
- Collaborative and version controlled work environments (e.g. gitlab, github, etc.)
- Python 3, C/C++
- Unix/Linux and any related scripting shell
- Differential calculus, vector analysis, combinatorial optimization (at least the basic principles thereof), convexity in optimization problems
-
Additional Knowledge and Professional Experience
- Fluent written and spoken English
- Knowledge of the concept of distributed computing, parallelization paradigms and COMPSs framework
- Knowledge of the difference between an AI instance and an ML model
-
Competences
- Good communication and presentation skills.
- Ability to work both independently and within a team.
Conditions
- The position will be located at BSC within the Computer Sciences Department
- We offer a full-time contract (37.5h/week), a good working environment, a highly stimulating environment with state-of-the-art infrastructure, flexible working hours, extensive training plan, restaurant tickets.
- Duration: Temporary
- Holidays: 23 paid vacation days plus 24th and 31st of December per our collective agreement
- Salary: we offer a competitive salary commensurate with the qualifications and experience of the candidate and according to the cost of living in Barcelona
- Starting date: 16/09/2024
Applications procedure and process
All applications must be made through BSC website and contain:
A full CV in English including contact details
A Cover Letter with a statement of interest in English, including two contacts for further references - Applications without this document will not be considered
In accordance with the OTM-R principles, a gender-balanced recruitment panel is formed for every vacancy at the beginning of the process. After reviewing the content of the applications, the panel will start the interviews, with at least one technical and one administrative interview. A profile questionnaire as well as a technical exercise may be required during the process.
The panel will make a final decision and all candidates who had contacts with them will receive a feedback with details on the acceptance or rejection of their profile.
At BSC we are seeking continuous improvement in our recruitment processes, for any suggestions or feedback/complaints about our Recruitment Processes, please contact recruitment [at] bsc [dot] es.
For more information follow this link
In accordance with the OTM-R principles, a gender-balanced recruitment panel is formed for every vacancy at the beginning of the process. After reviewing the content of the applications, the panel will start the interviews, with at least one technical and one administrative interview. A profile questionnaire as well as a technical exercise may be required during the process.
The panel will make a final decision and all candidates who had contacts with them will receive a feedback with details on the acceptance or rejection of their profile.
At BSC we are seeking continuous improvement in our recruitment processes, for any suggestions or feedback/complaints about our Recruitment Processes, please contact recruitment [at] bsc [dot] es.
For more information follow this link
Deadline
The vacancy will remain open until a suitable candidate has been hired. Applications will be regularly reviewed and potential candidates will be contacted.
OTM-R principles for selection processes
BSC-CNS is committed to the principles of the Code of Conduct for the Recruitment of Researchers of the European Commission and the Open, Transparent and Merit-based Recruitment principles (OTM-R). This is applied for any potential candidate in all our processes, for example by creating gender-balanced recruitment panels and recognizing career breaks etc.
BSC-CNS is an equal opportunity employer committed to diversity and inclusion. We are pleased to consider all qualified applicants for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, disability or any other basis protected by applicable state or local law.
For more information follow this link
BSC-CNS is an equal opportunity employer committed to diversity and inclusion. We are pleased to consider all qualified applicants for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, disability or any other basis protected by applicable state or local law.
For more information follow this link