An analysis of > 350,000 registered crimes over a 10 year span in Washington DC was conducted. This was accomplished using R, SQL and Apache Hadoop.
DC has long struggled with a history inner-city crime. The nickname 'Murder Capital' was bestowed on the city in the late 80's. In 1997, the Washington NBA basketball team even changed its name from the Bullets to the Wizards. The aim of this investigation is to better understand what contributing factors influence violent crime in specific. 9 unique datasets derived from opendata.gov were merged together into 1. A multitude of models were created. These models differentiate violent and non-violent crime with different degrees of success and are detailed in the write-up. Some data imputation was required to acheive this, including the creation of a violent/non-violent variable. Understanding what predicts violent crime enables police to allocate resources more effectively. Furthermore, preventative measures can be targeted more specifically.
Write-up and Modeling code are available.