Data Benchmark Collection of Patients with Malaria for Machine Learning: a study in Hadhramout- Yemen

Abstract

Studies on data collection of health and medicine are lacking; especially in endemic diseases. This study aimed to create and publish a data benchmark and make it available for researchers. Malaria is one of the endemic and infectious diseases consistently in many countries. must attention. Such diseases need high attention to diagnose, predict, and control by physicians and stakeholders. The initial step of data mining and data analysis is data collection to draw knowledge of this data and speed in the elicitation information. Using data mining techniques can enable one to build predictive models. About a thousand cases of finally diagnosed malaria patients have been collected under the direct supervision of specialized doctors and preprocessed for future use. We reported the percentage of patients with data recorded for 40 attributes and selected 27 signs and symptoms. Data quality was assessed based on statistical scores. Later, Basic analysis is presented in this paper too, employing common statistics metrics and data mining as well. This study uses the SPSS tool for statistical analysis and WEKA as a tool for data mining.