A very common workflow is to index some data based on its embeddings and then given a new query embedding retrieve the most similar examples with k-Nearest Neighbor search. For example, you can imagine embedding a large collection of papers by their abstracts and then given a new paper of interest retrieve the most similar papers to it.
TLDR in my experience it ~always works better to use an SVM instead of kNN, if you can afford the slight computational hit
We observed that generally the embedding representation is very rich and information dense. For example, reducing the dimensionality of the inputs using SVD or PCA, even by 10%, generally results in worse downstream performance on specific tasks.
The present study investigated mosquito species composition and phenotypic insecticide resistance profile to support decision-making in vector control in the Republic of Djibouti at the Horn of Africa. Adult mosquitoes were collected between December 2016 and December 2017 across 20 sentinel sites established in the 6 regions of the country using both Centers for Disease Control (CDC) miniature light traps and pyrethrum spray catches (PSC). Female mosquitoes were kept aside, for morphological identification to species by an expert entomologist using appropriate taxonomic keys by Gillies & Coetzee and Glick. Bioassays were also conducted in An. stephensi from Djibouti-ville against nine insecticides used in public health. A total number of 12,538 host-seeking mosquitoes belonging to four genera (Anopheles, Culex, Aedes, Uranotaenia) comprising 12 species were collected. Among these, A. gambiae S.L. and A. stephensi were the two major malaria vectors identified while secondary malaria vectors such as A. nili somalicus, A. dthali and A. azaniae were also collected. Culex quinquefasciatus was the most abundant mosquito species in the 6 regions. WHO susceptibility tests performed on A. stephensi population from Djibouti-ville showed resistance to pyrethroids, organophosphates, carbamates and DDT.