On Scene Injury Severity Prediction (OSISP) Algorithm Improvement Using Data Science
The objective was to enhance an existing OSISP algorithm in four dimensions. These dimensions are performance, validity, usability and generalizability. The approach is based on the use of big data tools and technologies. Our previous OSISP studies are based on homogeneous data from a single source (STRADA). In this project, a systematic exploration of different data analysis and machinelearning approaches has been carried out. Due to the heterogeneous data sources (STRADA and NASS-CDS), a suite of data mapping and encoding schemes was defined and implemented using efficient data storage and retrieval protocols for improved performance.