基于商空间理论的道路交通事故分析

Puts forward analysis of past accidents by granular computing methods. First, the theoretical framework of quotient space granular computing, the characteristic attributes can be gotten by the analysis of the accident over the years the elements; Secondly, the statistical algorithm makes the decision-making attributes by analyzing the past incidents data. The characteristic and decision-making attributes are combined to forecast the traffic incident by constructive machine learning methods. The establishment of an accident prediction model orders to achieve more accurate forecasting and analysis of the accident. Experiments prove that based on the quotient space granular computing theory this prediction method to accident is effective.