We seek outstanding candidates to work on a two-year project to
develop machine learning tools to predict NHS hospital attendance at
St Mary’s Hospital in Paddington, London.
We seek an exceptional and technically proficient Research Assistant
or Associate to utilise the current state-of-the-art techniques from
machine learning and statistics to investigate:
- To what degree is demand fundamentally predictable from an
information theoretic and entropy view.
- What statistical architecture best captures the underlying temporal
process (e.g ARIMAs, Smoothing, Long short term memory/GRU recurrent
nets, seq2seq architectures etc.)
- What quantitative accuracy can be expected from forecasts and how
this degrades with the level of aggregation and time horizon of the
forecast,
- By collecting accessory data, and including random effects, what are
the underlying factors contributing to changes in demand,
- Can detailed calendar information such as school holidays, weather,
and public events be used to leverage predictive accuracy
- Is it possible to characterise the frequency and occurrence of
anomalous surges of excessive demand
- Can our framework be applied efficiently in near-real time settings?
Candidates with a machine learning background and experience with time
series are encouraged to apply.
For informal inquiries contact Samir Bhatt ([log in to unmask]) or
Seth Flaxman ([log in to unmask])
https://www.jobs.ac.uk/job/BMY128/research-assistant-or-associate
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|