The purpose of this demo is to demonstrate the concept of word embeddings using textual data. Word embeddings are a form of language modeling using natural language processing and machine learning techniques in order to map words from a selected vocabulary to a vector of numbers.

The intention of this project is to perform an analysis on a representative sample of death records from public data available for purchase from the Washington State Department of Health prior to performing a analysis based on the full set of data for all United States records from 2003-2017. The focus of the project will be examining records pertaining to drug overdose deaths so we will limit our analysis to a subset of ICD-10 codes.

2.4 million records from Baltimore City speed cameras from 2013 to the present.