Main area of interest are:
- Artificial Intelligence
- Applied statistics
- Machine learning
- Data mining
- Natural language processing
Dr Mirkes elaborated a new and universal framework to minimise arbitrary sub-quadratic error potentials in machine learning and developed new machine learning methods which achieve orders of magnitude with faster computational performance than corresponding state-of-the-art methods.
The new powerful version of Supervised PCA was developed. This version allow significantly reduce dimension of space used to solve specific classification problem.
Several projects was performed with College of Life Science: with TARN in collaboration with Professor Timothy Coats and Dr Damian Roland and with Professor Thomas Yates, Dr Alex Rowlands and Dr Petra Jones.
Several collaborative project with Professor Stanislav S. Piletsky from chemistry department. There were several successful industrial projects with Weatherford Ltd. Bellrock Group SPRINT and other.
Fehrman, E., Egan, V., Gorban, A.N., Levesley, J., Mirkes, E.M. and Muhammad, A.K., 2019. Personality Traits and Drug Consumption: A story told by data. Springer.
Manso, A.S., Chai, M.H., Atack, J.M., Furi, L., Croix, M.D.S., Haigh, R., Trappetti, C., Ogunniyi, A.D., Shewell, L.K., Boitano, M. and Clark, T.A., 2014. A random six-phase switch regulates pneumococcal virulence via global epigenetic changes. Nature communications, 5(1), pp.1-9.
Mirkes, E.M., Coats, T.J., Levesley, J. and Gorban, A.N., 2016. Handling missing data in large healthcare dataset: A case study of unknown trauma outcomes. Computers in biology and medicine, 75, pp.203-216.
Gorban, A.N., Mirkes, E.M. and Zinovyev, A., 2018, July. Data analysis with arbitrary error measures approximated by piece-wise quadratic PQSQ functions. In 2018 International Joint Conference on Neural Networks (IJCNN) (pp. 1-8). IEEE.
Rowlands, A.V., Mirkes, E.M., Yates, T., Clemes, S., Davies, M., Khunti, K. and Edwardson, C.L., 2017. Accelerometer-assessed physical activity in epidemiology: are monitors equivalent?