Internship Title
Internship – Exploration of statistics and Machine Learning techniques for nuclear reaction
data curation
Duration of Internship
3 to 12 months
Organizational Setting
Department: Nuclear Sciences and Applications (NA)
Division: Physical and Chemical Sciences (NAPC)
Section: Nuclear Data Section (NDS)
Unit: Nuclear Data Development Unit (NDDU) and Nuclear Data Services Unit (NDSU)
Main Purpose
The main purpose of the internship is to prototype modern methods and software tools for
the nuclear data development.
During the duration of the internship with the Nuclear Data Section, the intern will explore the
potential of statistical algorithms and Machine Learning (ML) techniques as tools to assist in
the curation of nuclear reaction data from databases, such as identifying anomalies and
enhancing datasets with meta-data through statistical imputation methods or natural
language processing of relevant literature. The intern will be also getting acquainted with
general principles of:
• Nuclear data as a foundation to pure and applied nuclear science and engineering
• Statistical modelling in the context of nuclear data
• Data handling with experimental nuclear reaction databases such as EXFOR as well
as evaluated nuclear data libraries such as FENDL, TENDL, Neutron Data Standards
• Tools to interact with and update nuclear data
Tasks / Key Results Expected
Depending on progress, interest and priorities, a selection of the following:
• Create prototypes for the validation and augmentation of nuclear databases, such as
o Creating scripts to flag and visualize anomalies in databases
o Designing scripts to identify relevant literature with nuclear data from online
literatures
o Integrating these scripts into evaluation pipelines for more robust data
evaluations
o Constructing scripts with Natural Language Processing to improve database
quality or generate database entries automatically
Knowledge, Skills and Abilities
• Excellent programming skills in Python
• Basic knowledge of statistics and machine learning techniques.
• A good foundation in linear algebra and numerical optimization is an asset.
• Experience with databases with scientific data is an asset.
• Good communication and teamwork skills; capable of working independently and
proactively.
Qualifications and Experience
• University degree in Computer Science, Data Science, Physical Sciences,
Mathematics, Engineering, or a related discipline.
• Experience in the application of Python to solve problems involving or related to
scientific data.
Potential Institutions/Organizations that can be reach out to in order to attract potential
applicants
• Technical Universities.
• Research Institutes.
Internships
The IAEA accepts a limited number of interns each year. The internships are awarded to
persons studying towards a university degree or who have recently received a degree (see
Internship web pages for further details).
The purpose of the programme is:
• To provide interns with the opportunity to gain practical work experience in line with
their studies or interests, and expose them to the work of the IAEA and the United
National as a whole.
• To benefit the IAEA's programmes through the assistance of qualified students
specialized in various professional fields.
• The duration of an internship is normally not less than three months and not more
than one year.
Applicant Eligibility
• Candidates must be a minimum of 20 years of age and have completed at least three
years of full-time studies at a university or equivalent institution towards the
completion of a first degree.
• Candidates may apply up to one year after the completion of a bachelor's, master's
or doctorate degree.
• Candidates must not have previously participated in the IAEA's internship
programme.
• Excellent written and spoken English essential; fluency in any other IAEA official
language (Arabic, Chinese, French, Russian) an asset.
• Candidates must attach two signed letters of recommendation to their application.
