论文信息 - Analyzing Accident Prone Regions by Clustering

Analyzing Accident Prone Regions by Clustering

Traffic accidents and injuries related to them have unfortunately become a daily incident for the people of Bangladesh and this is particularly true for people living in Dhaka City. This paper aims to identify the most hazardous regions for such incidents within the Dhaka Metropolitan Region as well as assess their influences. This research effort collects accident related data from the Accident Research Institute (ARI) at Bangladesh University of Engineering and Technology (BUET), Dhaka. This paper utilizes the k-means clustering and expectation maximization method to cluster related incidents together.

[1] Yetis Sazi Murat. Fuzzy Clustering Approach for Accident Black Spot Centers Determination , 2012 .

[2] H. Hartley. Maximum Likelihood Estimation from Incomplete Data , 1958 .

[3] Durga Toshniwal,et al. A data mining framework to analyze road accident data , 2015, Journal of Big Data.

[4] Abdülhamit Subaşi,et al. Comparison of clustering techniques for traffic accident detection , 2015 .

[5] Greg Hamerly,et al. Alternatives to the k-means algorithm that find better clusterings , 2002, CIKM '02.

[6] Kamruzzaman,et al. Analysis of traffic injury severity in a mega city of a developing country , 2013 .

[7] Mazharul Hoque,et al. Road safety in Bangladesh and some recent advances , 2007 .

[8] Bernadette A. Thomas,et al. Global, regional, and national age–sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013 , 2015, The Lancet.

[9] Maya R. Gupta,et al. EM Demystified: An Expectation-Maximization Tutorial , 2010 .