Deep Data Insight Master Person Index
In current data driven business environment Master Person Index (MPI) applications are becoming more and more integral to all industries ranging from healthcare to education where information of the same person is recorded in different places without any reference linking these together. Deep Data Insight MPI system analyzes data sources, learns and recalls previously seen records, creates and maintains a master index and references to each original data source. In addition to these, the system is also equipped with end user assistance through text based search and facial recognition allowing the end user to query the system using full/partial information of the person or using images of the person. This system provides fast, effective and secure MPI solutions for any industry including Healthcare, Insurance, Supply Chain and Education. In these industries, this system has the impact in increasing the efficiency and effectiveness of person/entity coordination, management, analytics, etc.
Time Series Predictive Model for Dengue Cases in Sri Lanka
Dengue disease has been identified as a rapidly developing pandemic-prone viral disease in many countries in the world. In recent decades, dengue incidence has dramatically spread in worldwide. Severe dengue is a leading of serious illness and presently, it affects Asian and Latin American countries. Since particular treatment has not been identified for dengue or severe dengue, early detection is very important to reduce the fatality rates by accessing proper medical care. This study is mainly focus on developing time series predictive model to forecast dengue incidence in future months in Sri Lanka. Mainly, monthly dengue cases reported in Sri Lanka during January 2010 to September 2017 has been used for developing time series model. Here, SARIMA (2, 1, 2) (0, 0, 1)4 model has been selected as the most appropriate model based on the AIC value. Also, the model can be used for predicting number of dengue cases in Sri Lanka if the observations of time series do not indicate unusual dengue incidence in future months. The validation criterion for the fitted model has been satisfied and accuracy of the fitted model was checked using measurement of errors (i.e. RMSE, MAPE etc.) which have indicated as satisfactorily small.