Problem solving in mathematics, statistics and data science is inspiring and enjoyable, but are achievements from these fields any use in the so-called 'real world'?
The work often turns out useful and essential, but typically in a less concrete manner than say the work of a scientists or a physician. David Hilbert, in his now historical address to scientists and physicians, put it this way:
"The instrument that mediates between theory and practice, between thought and observation, is mathematics; it builds the connecting bridge and makes it stronger and stronger. Thus it happens that our entire present-day culture, insofar as it rests on intellectual insight into and harnessing of nature, is founded on mathematics"
Almost a century after Hilbert's words, the mathematical fundations of sciences and social sciences, and the evidence based approach in medicine are often being taken for granted. In the 21st century we are facing complex big data sets with unknown structures and and ongoing discourses about objective and subjective notions of risk and uncertainty.
Probability and statistics are mathematical disciplines for modelling and analysing theoretical and practical questions arising from this. In conjunction with data, they can help understanding real-world problems and make a difference.
Researchers in Statistics at Warwick are developing and utilizing modern statistics, mathematics and data science to solve practical problems such as:
- Applying computational statistics algorithms to analyse urban air quality
- Using quantile regression on brain image data to model progression of Alzheimer's disease
- Discovering which genes' activity can discriminate between diseased and healthy patients
- Analysing sports data to describe players and teams performance
- Modelling and detecting asset price bubbles while they are happening and before they burst
- Modelling infectious diseases and identifying localized outbreaks
- Developing a fast algorithm through probabilistic modeling for compression of sound data
- Automatically diagnosing diseases with large-scale image data Utilizing crime data for crime prevention and optimal allocation of police resources
- Predicting the outcome of elections based on exit poll data
- Computed Tomography validation of complex structures in Additive Layer Manufacturing
Probability of containment for multitype branching process models for emerging epidemics
Non-stationary statistical modeling and inference for circadian oscillations for research in cancer chronotherapy
Bayesian Models of Category-Specific Emotional Brain Responses
Decision focused inference on Networked Proabilistic Systems: with applications to food security
Rotationally invariant statistics for examining the evidence from the pores in fingerprints
Dynamic Uncertainty Handling for Coherent Decision Making in Nuclear Emergency Response
Study of Key Interventions into Terrorism using Bayesian Networks
Assessing the risk of subsequent tonic-clonic seizures in patients with a history of simple or complex partial seizures
Multidimensional Markov-functional Interest Rate Models
Prospect Theory, Liquidation and the Disposition Effect
Dynamic Bradley-Terry modelling of sports tournaments