Machine learning algorithms for detecting dangerous mosquito species (BAGNALL_U17DTP)

Lead Research Organisation: University of East Anglia
Department Name: Graduate Office

Abstract

One of the most significant challenges in the combat against vector based infectious diseases such as malaria, dengue and zika is the lack of extensive information as to the locations and dynamics of the populations of the disease disease-bearing mosquitoes. If we were able to automatically detect the presence of different species of mosquitos it could transform our ability to tackle these diseases. Detecting sensors could be used to track the effectiveness of preventative measures, act as early warning devices for population growth in dangerous species and provide long term data on changes in abundance.

Unfortunately, insect surveillance is time consuming and expensive with current technology. However, recent advances in sensor technology [1] have indicated that it is possible to differentiate species by their sound in flight. Researchers at the University of California, Riverside (UCR), have built inexpensive devices that can detect flying insects passing through the sensor and use photonics to recreate the sound of the insect. This project involves extending recent advances in time series classification [2] and speech processing [3] to develop algorithms that can accurately classify the species of mosquito based on its sound. There will be opportunities of placements in California and Kenya to collaborate with researchers working in this area.

[1] Y. Chen, A. Why, G. Batista, A. Mafra-Neto, E. Keogh. Flying Insect Classification with Inexpensive Sensors. Journal of Insect Behavior: 27(5). 657-677, 2014
[2] A. Bagnall, J. Lines, J. Hills and A. Bostrom. Time series classification with COTE. IEEE Trans. KDE: 27 (9). 2522-2535, 2015.
[3] P. Harding and B. Milner. Reconstruction-based speech enhancement from robust acoustic features. Speech Comm.: 75. 62-75, 2

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
BB/M011216/1 01/10/2015 30/09/2023
1916139 Studentship BB/M011216/1 01/10/2017 30/09/2021 Michael James Flynn