Disease Trajectory Mining Using Network Analysis

Abstract

Pakistan is struggling to provide good and on time healthcare facilities to its people.Situation become worse when the number of patients exceeds the capacity of available hospitals and healthcare facilities. With the help of current technology and data availability, we can analyse and forecast particular disease occurrence in the form of disease locations.This information is useful for healthcare sector and policy makers to make better decisions.This research analyse and forecast SARS and cholera disease locations for different cities of Punjab province.To conduct the study deep learning models LSTM(long short term memory) and Transformer models are used for forecasting dates and locations of SARS and cholera disease using 2016’s year data set. We believe that techniques used in this study to forecasting SARS and cholera disease dates and locations can be extended to other diseases as well.This will play an important role in improving the public health situation of Pakistan.Due to the data set availability we target different cities of Punjab province, however it can be extend to other province cities as well.