Impact through Data Science and AI

Data Science and Artificial Intelligence research group

We perform fundamental and applied research in the area of Data Science, Machine Learning and Artificial Intelligence (AI). This ranges from better understanding the neuroscience of the human brain, to advancing the state-of-the-art in Machine Learning and AI, and solving real-world problems across a number of application domains.

The majority of our work is cross-disciplinary, collaborating with domain experts within and outside the university. We believe in solving problems that matter and have a keen interest in collaborating with external organisations from both industry and academia.

We work confidently with a wide range of application domains, including health, finance and smart cities. Our technical expertise lies in doing clever things with data, such as machine learning, deep learning, data mining, forecasting, computational intelligence, search & optimisation, simulation modelling, complex networks and human-machine networks.

Our projects:

We solve real-world problems and advance the state-of-the-art in data science and AI. A selection of our portfolio can be found below.

Projects

Efficient spinal assessment through AI

Lower back pain is the world’s greatest cause of population days lost to disability and most of this is attributable to pain that has become chronic. Quantitative Fluoroscopy (QF) is the gold standard measure of the segmental biomechanics of this condition but is limited in its range to patients whose bone images are of high quality, which unfortunately excludes many older, obese, osteoporotic or scoliotic patients.

This project is the first step towards widening the ranges of age, morphology and complaint type that can be examined using QF through protocol improvements via applying Artificial Intelligence (AI) techniques to the registration and tracking of the vertebrae, making the assessment more efficient. This will enable care to be informed by objective and individualised assessment at crucial stages of the condition.

Funded by the Medical Research Council (MRC), this project started in August 2022.

Novel AI solutions for law enforcement

The purpose of this project is to develop novel AI solutions for law enforcement agencies to automatically assess the value of evidential material to assist in the investigative process and improve the detection of crime, extend the capability to derive risk and predictive modelling to shape resource planning and deployment.

Funded by InnovateUK, this Knowledge Transfer Partnership (KTP) project will start in February 2023.

FoodMAPP

Foundational redesign of our food supply systems is required, with urgency, to address the interconnected challenges of climate change, food supply resilience, food waste and supporting the next generation of producers and operators with a fair income.

FoodMAPP will address these urgent challenges by developing a responsive smart platform providing transparency of local food supply enabling more sustainable consumption practices, reducing food waste and supporting business development including small scale family farming and providing retail opportunities for food operators.

Funded by EC Horizon, this project will start in March 2023.

Market-driven chatbots

Shoeprint analysis to aid law enforcement

Mobile app for automated thyroid drug titration

A question mark formed out of pills on a yellow background

Royal Bournemouth Hospital (RBH) have over 1,000 thyroid patient contacts per year. They are primarily treating patients with hyperthyroidism/thyrotoxicosis, the condition that occurs due to excessive production of thyroid hormone by the thyroid gland. This involves holistic management of the patient but invariably the titration of anti-thyroid drugs – usually Carbimazole.

The dosing of Carbimazole is high to start with and is down-titrated often to zero over a period of 6-18 months. The dose adjustment is based on experience and requires 4-6 weekly assessment of thyroid function via a blood test. A healthcare professional (HCP, doctor or nurse) then rings or writes to the patient to alter the dose of the medication. This often introduces unnecessary delays in the process (1-2 weeks for the patient to receive the letter) and is very time intensive for the HCPs. Unlocking some of the time lost while performing these mundane tasks could lead to more balanced workloads of time-stretched HCPs and increased productivity, ultimately improving quality of care.

Funded through local HEIF funding, this project finished July 2022.

Events and activity

Deep Learning Winter School

We are hosting the 8th International School on Deep Learning, 16 - 20th January 2023.

Information and registration details can be found on the Deep Learn 2023 website

Innovations in Digital Health and Wellbeing

In collaboration with Frontiers in Big Data, (CiteScore: 2.7), Dr Vegard Engen is co-editing a new article collection (Research Topic) titled: “Innovations in Digital Health and Wellbeing”.

The aim of the Research Topic is to bring together the latest quality articles from researchers working in the area of Digital Health and Wellbeing, focused on either the operational aspects of patient care (e.g. through supply and demand forecasting, discharge prediction and estimating risk of readmissions), or improving treatment and patient care, both on-premise and home or virtual care (using state-of-the-art technologies involving, e.g., Big Data Processing, Visual Analytics, Data Science / Machine Learning, IoT or HCI methods).

Find out more on the Frontiers website

Key publications

Balaguer-Ballester, E., Nogueira, R., Abolafia, J.M., Moreno-Bote, R. Sanchez-Vives, M.V. 2020. Representation of Foreseeable Choice Outcomes in Orbitofrontal Cortex Triplet-wise Interactions. Plos Computational Biology, 16(6): e1007862.

Tabas, A., Andermann, M. (shared first authorship), Schuberth, V., Riedel, H., Balaguer-Ballester, E., Rupp, A. (shared senior authorship). 2019. Modelling and MEG evidence of early consonance processing in auditory cortex. Plos Computational Biology, 15(2): e1006820.

Nogueira, R., Abolafia, J.M., Drugowitsch, J., Balaguer-Ballester, E., Sanchez-Vives, M.V., Moreno-Bote, R. 2017. Lateral orbitofrontal cortex anticipates choices and integrates prior with current information. Nature Communications, 8. DOI: 10.1038/ncomms14823.

Lapish, C. and Balaguer-Ballester, E. (shared first authorship), Phillips, A, Seamans, J. and Durstewitz, D. 2015. Amphetamine Bidirectionally alters Prefrontal Cortex Attractor Dynamics during Working Memory. The Journal of Neuroscience 35(28): 10172-10187.

Hyman, J., Ma, L., Balaguer-Ballester, E., Durstewitz, D., Seamans, J. 2012. Contextual encoding by ensembles of medial prefrontal cortex neurons. Proceedings of the National Academy of Sciences USA (PNAS), 109 (13)5086-5091.

Nandakumar, V., Swain, I., Taylor, P., Merson, E. and Budka, M., 2022. SmartStim: A Recurrent Neural Network Assisted Adaptive Functional Electrical Stimulation for Walking. Current Directions in Biomedical Engineering, 8 (3), 41-43.

Budka, M., Ashraf, A.W.U., Bennett, M., Neville, S. and Mackrill, A., 2021. Deep multilabel CNN for forensic footwear impression descriptor identification. Applied Soft Computing, 109.

Budka, M., Bennett, M.R., Reynolds, S.C., Barefoot, S., Reel, S., Reidy, S. and Walker, J., 2021. Sexing white 2D footprints using convolutional neural networks. PLoS ONE, 16 (8 August).

Wahid-Ul-Ashraf, A., Budka, M. and Musial, K., 2019. How to predict social relationships — Physics-inspired approach to link prediction. Physica A: Statistical Mechanics and its Applications, 523, 1110-1129.

Salvador, M.M., Budka, M. and Gabrys, B., 2019. Automatic Composition and Optimization of Multicomponent Predictive Systems With an Extended Auto-WEKA. IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING, 16 (2), 946-959.

Almilaji, O., Engen, V., Snook, J. and Docherty, S., 2022. The development of a web-based application to predict the risk of gastrointestinal cancer in iron deficiency anaemia; the IDIOM app. Digital, Special Edition of Intelligent Digital Health Interventions, 2 (1), 104-119.

Almilaji, O., Engen, V., Thomas, P. and Snook, J., 2021. THE DEVELOPMENT OF A WEB-BASED APPLICATION TO PREDICT THE RISK OF GI CANCER IN IDA. GUT, 70, A37-A38.

Kyriazis, D., Engen, V. et al., 2019. The CrowdHEALTH project and the hollistic health records: Collective wisdom driving public health policies. Acta Informatica Medica, 27 (5), 369-373.

Jamil, W. and Bouchachia, A., 2022. Iterative ridge regression using the aggregating algorithm. Pattern Recognition Letters, 158, 34-41.

Pedrosa, J., Bouchachia, H. et al., 2021. LNDb challenge on automatic lung cancer patient management. Medical Image Analysis, 70.

Kyamakya, K., Al-Machot, F., Mosa, A.H., Bouchachia, H., Chedjou, J.C. and Bagula, A., 2021. Emotion and stress recognition related sensors and machine learning technologies. Sensors, 21 (7).

Anwary, A.R., Cetinkaya, D., Vassallo, M. and Bouchachia, H., 2021. Smart-Cover: A real time sitting posture monitoring system. Sensors and Actuators, A: Physical, 317.

Pohl, D., Bouchachia, A. and Hellwagner, H., 2020. Active Online Learning for Social Media Analysis to Support Crisis Management. IEEE Transactions on Knowledge and Data Engineering, 32 (8), 1445-1458.

Majumdar, A., Bakirov, R., Hodges, D., Scott, S. and Rees, T., 2022. Machine Learning for Understanding and Predicting Injuries in Football. Sports Medicine - Open, 8 (1).

Bakirov, R., Fay, D. and Gabrys, B., 2021. Automated adaptation strategies for stream learning. Machine Learning, 110 (6), 1429-1462.

Bakirov, R., Gabrys, B. and Fay, D., 2016. Augmenting adaptation with retrospective model correction for non-stationary regression problems. Proceedings of the International Joint Conference on Neural Networks, 2016-October, 771-779.

Bhullar, G., Khullar, A., Kumar, A., Sharma, A., Pannu, H.S. and Malhi, A., 2022. Time series sentiment analysis (SA) of relief operations using social media (SM) platform for efficient resource management. International Journal of Disaster Risk Reduction, 75.

Singh, A., Pannu, H.S. and Malhi, A., 2022. Explainable Information Retrieval using Deep Learning for Medical images. Computer Science and Information Systems, 19 (1), 277-307.

Arora, P., Mishra, A. and Malhi, A., 2022. Machine learning Ensemble for the Parkinson’s disease using protein sequences. Multimedia Tools and Applications.

Huotari, M., Arora, S., Malhi, A. and Främling, K., 2021. Comparing seven methods for state-of-health time series prediction for the lithium-ion battery packs of forklifts. Applied Soft Computing, 111.

Malhi, A.K., Batra, S. and Pannu, H.S., 2019. An Efficient Privacy Preserving Authentication Scheme for Vehicular Communications. Wireless Personal Communications, 106, 487-503.

Tsimperidis, I., Rostami, S., Wilson, K. and Katos, V., 2021. User Attribution Through Keystroke Dynamics-Based Author Age Estimation. Lecture Notes in Networks and Systems, 180, 47-61.

Stubbs, R., Wilson, K. and Rostami, S., 2020. Hyper-parameter Optimisation by Restrained Stochastic Hill Climbing. Advances in Intelligent Systems and Computing, 1043, 189-200.

Wilson, K. and Rostami, S., 2019. On the integrity of performance comparison for evolutionary multi-objective optimisation algorithms. Advances in Intelligent Systems and Computing, 840, 3-15.

Mavengere, N.B., Henriksen-Bulmer, J., Passmore, D., Mayes, H., Fakorede, O., Coles, M. and Atfield-Cutts, S., 2021. Applying innovative technologies and practices in the rapid shift to remote learning. Communications of the Association for Information Systems, 48, 185-195.

Atfield-Cutts, S., Ollis, G., Coles, M. and Mayes, H., 2016. Blended Feedback II: Video feedback for individual students is the norm, on an undergraduate computer programming unit. In: 27th Annual Workshop of the Psychology of Programming Interest Group - PPIG 2016 7-10 September 2016 St. Catharine's College, University of Cambridge, UK. Psychology of Programming Interest Group (PPIG 2016).

Pandey, H., Aslam, M.S., Tiwari, P. and Band, S.S., 2022. Observer–Based Control for a New Stochastic Maximum Power Point tracking for Photovoltaic Systems With Networked Control System. IEEE Transactions on Fuzzy Systems.

Yadav, S.K., Agarwal, A., Kumar, A., Tiwari, K., Pandey, H.M. and Akbar, S.A., 2022. YogNet: A two-stream network for realtime multiperson yoga action recognition and posture correction. Knowledge-Based Systems, 250.

He, L., Guo, C., Tiwari, P., Su, R., Pandey, H.M. and Dang, W., 2022. DepNet: An automated industrial intelligent system using deep learning for video-based depression analysis. International Journal of Intelligent Systems, 37 (7), 3815-3835.

Das, S.R., Hota, A.P., Pandey, H.M. and Sahoo, B.M., 2022. Industrial power quality enhancement using fuzzy logic based photovoltaic integrated with three phase shunt hybrid active filter and adaptive controller. Applied Soft Computing, 121.

Yadav, S., Kera, S.B., Gonela, R.V., Tiwari, K., Pandey, H. and Akbar, S.A., 2022. TBAC: Transformers Based Attention Consensus for Human Activity Recognition. In: IEEE WCCI 2022 International Joint Conference on Neural Networks (IJCNN 2022) 18-23 July 2022 University Padua Italy.