Mining Road Traffic Accident Data of N5 National (Dhaka-Banglabandha) Highway in Bangladesh
Noor, Nishat Tasnim
Hoque, Mohammed Imranul
MetadataShow full item record
Road traffic accidents have now become a national issue as every year 2 to 3 percent of the GDP are wasting by traffic accidents in Bangladesh. A series of road accidents killed and injured a good number of people every year. According to police report, 2,513 people were killed in 2017 while the number was 2,463 in 2016. In order to prevent road accidents in the country, the government took several initiatives including no more than five hours driving on the highways. In this paper, we have analysed the road traffic accident data of national N5 (Dhaka- Banglabandha) highway of Bangladesh. The main objective of this research is to find the accident patterns and extract the rules from the historical accident data that helps us to prevent the road accidents. We have collected data of total 892 road accidents from the Accident Research Institute (ARI), Bangladesh University of Engineering Technology (BUET) formerly known as Accident Research Centre (ARC), which is a Center of Excellence for the advancement of safety research in Bangladesh. Initially, we have applied several data preprocessing techniques and extracted informative features from the data. Then, we have applied 10 most well-known and popular machine learning algorithms for data mining task on accident data and measured the performance of the classifiers.
- B.Sc Thesis/Project