Optimization and machine learning
[Closed]
Work description
Predictive maintenance problems are seldom studied under the perspective of integer optimization. Given their practical interest, data analytics and heuristics are widely used to address them, but the optimization (prescriptive) component is usually missing. This work will address the development of models and solution methods for this class of problems, with applications in situations where expensive, difficult to replace assets must be maintained for a long period. The scarcity of data typically available forces the use of models describing the physics underlying the aging process, as a complement to a data-driven approach. Our goal is to build models, mathematical optimization techniques for tackling them, complemented with data analytics techniques, aiming at a software tool able to solve practical instances. A case study on power transformers will be provided.
Academic Qualifications
Master's degree in computer science, operational research or related areas.
Preference factors
Fluency in English (spoken and written). Good mathematical modeling and programming skills.
Application Period
Since 05 Jan 2022 to 18 Jan 2022
[Closed]
Cluster / Centre
Industrial and Systems Engineering / Industrial Engineering and Management