Training the Classification Model

The classifier needs multiple samples collected from various insects to be able to differentiate signals generated by mosquitoes from signals generated by non-mosquitoes. We will buy bulks of commonly found insects instead of catching them to prevent damaged insects from being included in our experiments. Before training a classifier, appropriate features that differentiates mosquitoes from non-mosquitoes will be explored. The appropriate type of sensors will be determined based on the features considered. (e.g. acoustic signals coming from flight and wing beat such as frequency and rhythm, sounds generated by traps). Several classifiers will be trained and assessed to identify the most suitable for the task.