Applications are invited for a fully funded four-year PhD studentship investigating the use of novel machine learning methods to model and predict large-scale systems.
What is about: Large-scale systems are encountered frequently in engineering, biology systems, cyber-physical systems, finance and many other areas. As a practical example we can consider a large manufacturing plant. Large manufacturing plants require (i) to optimise their throughput; (ii) to detect any anomaly in their behaviour; (iii) to reduce waste (which is fundamental to improve their greenness). In relation to these purposes, plant performance/monitoring data, in the form of time-series, are collected at different multiple locations. What are the challenges? First, time series data are not temporally aligned and they are often intermittent. Second, due to the underlying interactions between the various components of these systems, the phenomena that produce these time series data interact and influence one another. Third, many component variables are manipulated and, therefore, they can change their value independently of the inputs of the component. These issues pose a challenge to traditional optimisation and machine learning methods that aim to use time-series data to accomplish those goals. The PhD student will design novel algorithms for multiple time-series extrinsic regression, which will go beyond the state-of-the-art approach (which nowadays consists in extracting features from the time-series and using traditional machine learning methods to accomplish those goals). In particular, the PhD student is expected to attack this problem from a different perspective, that is by considering time-series data as unstructured data and accounting for the causal interactions between the components of the system. Where: The School of Computer Science and Statistics at Trinity College Dublin is an innovative and energetic centre for academic study and research. Structured PhD training programme: This studentship is part of the SFI* centre for research training in AI <https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fcrt-ai.cs.ucc.ie%2Fprogramme.html&data=05%7C01%7Cuai%40engr.orst.edu%7Ce9a18f0f435d420c741208db0b6bcd5b%7Cce6d05e13c5e4d6287a84c4a2713c113%7C0%7C0%7C638116329857035305%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=5jJhg42zi3GGwDbSMS025oDRuux5s2582nzq1MIuU%2Fc%3D&reserved=0>. *PhD students within this programme will follow a structured PhD training programme that comprises four main elements: (i) Host-based research methods training; (ii) Supervisor-initiated research-specific training; (iii) CRT-organized training in Artificial Intelligence methods; and (iv) Work placements. Requirements: Applicants should hold at least a 2.1 honours undergraduate degree in Computer Science, Computer Engineering, Statistics, Mathematics, or a closely related area. Non-native English speakers are required to hold an IELTS certificate demonstrating at least a score of 6.5 overall with a score of not less than 6 in all parts. Funding: Student will receive a full scholarship to undertake a four year structured PhD programme. This scholarship comprises full payment of university fees for four years and a monthly tax-free stipend of €1,500 per month for four years. In addition, a budget for equipment, travel, and training is provided. Please contact me by email if interested: Alessio Benavoli, Trinity College Dublin (Ireland), alessio.benav...@tcd.ie
_______________________________________________ uai mailing list uai@engr.orst.edu https://it.engineering.oregonstate.edu/mailman/listinfo/uai
