Previous studies show that a trained partial least square regresser (PLSR) from near-infrared spectra classify laboratory and semi-field raised mosquitoes into less than or ≥ to seven days old with an average accuracy of 80%. This dissertation demonstrates that training models on near-infrared spectra (NIRS) using artificial neural network (ANN) as an architecture yields models with higher accuracies than training models using partial least squares (PLS) as an architecture. In addition, irrespective of the model architecture used, direct training of a binary classifier scores higher accuracy than training a regresser and interpreting it as a binary classifier. Furthermore, for the first time, this dissertation shows that training ANN models on autoencoded near-infrared spectra yields models that estimate parity status of wild mosquitoes with an accuracy of ≈93%, which is strong enough to support NIRS models as an alternative to ovary dissections. Results from this dissertation also show that there is no significant difference between spectra collected from semi-field raised and wild mosquitoes of the same species, supporting the on-going practice of training models on semi-field raised mosquitoes to estimate the age class in days of wild mosquitoes. Finally, the study shows that an ANN model trained on semi-field mosquitoes classifies wild mosquitoes into either less than or ≥ to seven days old with an average accuracy of 76%. In conclusion, the results in this dissertation strongly suggest the use of ANNs as a suitable architecture to train models that estimate parity status and age in days of wild mosquito vectors of infectious diseases. The results further suggest near-infrared spectroscopy as an appropriate alternative tool to estimate different parameters of mosquito vectors of infectious diseases.
|Advisor:||Corliss, George F., Bansal, Naveen|
|Commitee:||Kiware, Samson S., Bozdag, Serdar, Povinelli, Richard J.|
|Department:||Mathematics, Statistics & Computer Science|
|School Location:||United States -- Wisconsin|
|Source:||DAI-B 81/10(E), Dissertation Abstracts International|
|Subjects:||Artificial intelligence, Applied Mathematics, Computer science|
|Keywords:||Autoencoder and parity status, Clustering, Machine learning and data mining, Malaria and infectious diseases, Mosquitoes, Neural networks|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be