A Study on Traffic Accident Injury Severity Prediction Model Based on Public Data

This article describe harvest prediction model for the country or for the big region on the public available data. In the article are analysed impact of main fertilizers component and environmental variables to the grain harvest The aim of the article was to create regression model, which best describes grain harvest prediction on public (free) available data. Created final regression model explain 78% (R2) of the variation in the harvest result. Presented model show, that prediction accuracy significantly increase if environmental variables are added. Prediction accuracy (RMSE) of the final regression model was 3,89. All calculation was made on the example of the Germany.

Download Full-text

Traffic Accident Injury and Severity Prediction Using Machine Learning Algorithms

Lecture Notes in Electrical Engineering - ICCCE 2020 ◽

10.1007/978-981-15-7961-5_96 ◽

2020 ◽

pp. 1041-1048

Author(s):

Nithin Kashyap ◽

Hari Raksha K. Malali ◽

Koushik S. E ◽

Raju G ◽

T. H. Sreenivas

Keyword(s):

Machine Learning ◽

Traffic Accident ◽

Learning Algorithms ◽

Machine Learning Algorithms ◽

Severity Prediction ◽

Accident Injury

Download Full-text

(A175) Traffic Injury Severity Prediction by Algorithm of Automatic Crash Notification System

Prehospital and Disaster Medicine ◽

10.1017/s1049023x11001737 ◽

2011 ◽

Vol 26 (S1) ◽

pp. s49-s49

Author(s):

S.J. Wang ◽

H.Y. Choi

Keyword(s):

Medical Record ◽

Mobile Communication ◽

Traffic Accident ◽

Injury Severity ◽

University Hospital ◽

Insurance Companies ◽

Physical Injury ◽

Notification System ◽

Severity Prediction ◽

Medical Situation

IntroductionSince 2009 automatic crash notification system(ACNS) using event data recorder(EDR) and mobile communication have been developed for early detection of traffic accident and prediction of physical injury of victims for increase of survival rate via early medical treatment. For adequate prediction of injury, authors developed the guideline and algorithm from parameters related to accident and medical situation. Methods: Expert survey was done about the adequate parameters related to accident and medical situation. Medical record of traffic accident admission was analyzed in a trauma center of a university hospital in Seoul, Korea. Additionally epidemiology of traffic accident death in a region was done. Afterwards data of medical record was linked to data of traffic accident insurance companies.ResultsThe important parameters for prediction of physical injury of victims were as follows: Intercept, deltav, belt, age, intrus, sex, multiple, roll, ejection, narrow, height, weight, steering defect, track loc.ConclusionsPrediction of physical injury severity of victims on traffic accident spot and immediate transfer of related information to adequate medical institution by automatic mobile communication can help the traffic accident victims and upgrade the trauma care system of traffic accident.

Download Full-text