Sequential Minimal Optimization Algorithm with Support Vector Machine for Mosquito Larvae Identification

Efficient and fast mosquito larvae identification is of important due to the cost, effort and time taken during manual processes. This paper evaluates the use of the sequential minimal optimization algorithm (SMO) employed with support vector machine (SVM) to improve the identification process that takes into account mosquito larva images. All images of Aedes and Culex are transformed to binary values with the application of pre-processing steps. The comparison results are based on computational experiments of the linear kernel, polynomial kernel, and Gaussian Radial Basis Function (RBF) kernel settings. The findings of RBF offer a better performance compared to linear and polynomial.