MODELO PREDICTIVO EN LOS ACCIDENTES DE TRANSITO CON BASE EN DATA SCIENCE.

Abstract

The reason for road accidents has become one of the main causes of the greatest number of victims existing annually; This problem is especially concentrated in middle and low-income countries where there is no safe mobility plan or effective prevention measures related to traffic accidents to prevent deaths and injuries in society. In Ecuador and in the canton of Guayaquil specifically, this road accident problem is no stranger, so in this study a statistical analysis was carried out where the main sectors of the city were analyzed, taking into account factors such as time and day of the week of these incidents in the city, taking as a reference the base of accidents of the city, and based on the results, proposals and solutions can be identified in order to reduce the negative figures of this social problem, with the information provided by the entities that control the traffic in the canton, with the help of statistical software. Terms used in the theoretical study of the analyzed topic and the statistical methods used for this project are conceptualized, subsequently, through the 2016-2018 traffic accident database, a statistical analysis was carried out to know the current status of proportionality and occurrence of accidents in the canton, then the analysis of the probability of accidents was developed using the binary logistic regression model to obtain an estimate of whether or not there is a traffic accident for the year 2019 in the city.