INCREMENTAL TRANSDUCTIVE LEARNING APPROACHES To Schistosomiasis Vector Classification

摘要

The key issues pertaining to collection of epidemicrndisease data for our analysis purposes are that it is a labourrnintensive, time consuming and expensive process resulting inrnavailability of sparse sample data which we use to developrnprediction models. To address this sparse data issue, we presentrnthe novel Incremental Transductive methods to circumvent therndata collection process by applying previously acquired data tornprovide consistent, confidence-based labelling alternatives to fieldrnsurvey research. We investigated various reasoning approachesrnfor semi-supervised machine learning including Bayesian modelsrnfor labelling data. The results show that using the proposedrnmethods, we can label instances of data with a class of vectorrndensity at a high level of confidence. By applying the Liberalrnand Strict Training Approaches, we provide a labelling andrnclassification alternative to standalone algorithms. The methodsrnin this paper are components in the process of reducing thernproliferation of the Schistosomiasis disease and its effects.