Data science applied to the fight against traffic accidents

CategoryPersonal-Academic Project
SummaryThe project intend to show how the “life cycle” applies to data mining (DM) projects through applying a tentative exemplary case of circulation accidents
ToolsRStudio, Spreadsheets, BigQuery, Tableau
Repository Link
SkillsR, Git
TypeData Mining

We are requested to analyze the causes of road accidents in the United States, whether they are of human, material, or environmental origin. With this information, they hope to annually review the trends in this matter with the help of the model and adjust the intervention plan, whether through investment, campaigns, or training. They are also very interested in identifying specific states and cities where the intervention in road safety should be increased or modified, as well as any aspects related to infrastructure. Finally, we are asked to review the three-year time series to understand the evolution of road accidents.

This data mining project aims to explore the dataset, uncover hidden patterns, and in future if potentially possible, develop predictive models to identify factors contributing to severe accidents. The insights gained from this analysis will provide valuable information for road safety initiatives and support evidence-based decision-making in accident prevention strategies.

The primary analytical objective of this project is to gain insights into the factors that contribute to the severity of an accident and to define what constitutes a severe accident. By applying data mining techniques, we aim to uncover patterns and relationships within the dataset that can help us understand the key factors associated with severe accidents.

Based on the above and in summary, we will undertake the initial phases to design a data mining model that allows us to understand the following in an updated manner:

  1. The evolution of the time series of fatal accidents.
  1. Major causes of accidents.
  1. Incident volume by states and cities, including “black spots” in the road network.