The purpose of this paper is to propose a framework of building classification models to deal with the problem in predicting Schistosomiasis vector density. We aim to resolve this problem using remotely sensed satellite image extraction of environment feature values, in conjunction with data mining and machine learning approaches. In this paper we assert that there exists an intrinsic link between the density and distribution of the Schistosomiasis disease vector and the rate of infection of the disease in any given community; it is this link that the paper is focused to investigate. Using machine learning techniques, we want to accumulate the most significant amount of data possible to help with training the machine to classify snail density (SD) levels. We propose to use a novel cumulative training approach (CTA) as a way of increasing the accuracy when building our classification and prediction model.