Swansea University is a research‑led university that has been making a difference since 1920. The University community thrives on exploration and discovery and offers the right balance of excellent teaching and research, matched by an enviable quality of life. Our stunning waterfront campuses and multicultural community make us a desirable workplace for colleagues from around the world. Our reward and benefits, and ways of working enable those who join us to have enriching careers, matched by an excellent work‑life balance.
About The Role
The School of Management at Swansea University is seeking to appoint a part‑time Research Assistant to support the project “Explainable ML‑based Predictive Model for Thermal Comfort for Net Zero Buildings”. The project aims to develop a data‑efficient and explainable machine learning framework to predict thermal comfort and support smarter, lower‑carbon building management in the Swansea Bay City Region.
* Prepare and analyse indoor environmental quality and occupant comfort data, including sensor measurements and short comfort survey data.
* Clean, harmonise, and quality‑assure datasets; address missing values, sparse comfort labels, and class imbalance.
* Apply data augmentation methods such as SMOTE or GANs where appropriate.
* Develop, train, optimise, and evaluate compact machine learning models, including Gradient Boosting, Support Vector Regression, and shallow neural networks.
* Contribute to robust cross‑validation, Bayesian optimisation, reproducible model development, and SHAP‑based explainability to identify the influence of factors such as air temperature, mean radiant temperature, relative humidity, air speed, clothing insulation, and metabolic rate on thermal comfort.
* Support the development of a Building Management System‑ready advisory prototype, including clear visual outputs and practical recommendations on comfort‑preserving setpoints and schedules.
Contribute to project deliverables, technical reporting, academic outputs, stakeholder engagement, and knowledge exchange activities with regional partners, including building management, housing, and energy‑sector stakeholders.
This role would suit a motivated candidate with experience in Python‑based machine learning, real‑world sensor or time‑series data, and an interest in explainable AI, smart buildings, thermal comfort, or net‑zero innovation.
The post is fixed‑term, part‑time for 10 months and based at Swansea University’s Bay Campus.
#J-18808-Ljbffr