Name: Srikanth 
github ID: srikanth261
NYU ID: sn2495
PUI Extra credit project
Problem Description:
Predict the number of Violent Crimes and in specific , the number of murders in metro-areas using publicly available crime and census data with the help of a data/statistical model 
Data:
Data will be used primarily from the crime statistics reported on the city data website and also the US census data. 
The crime data can be scraped from http://www.city-data.com/crime/
The US census information is available at https://factfinder.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t
Analysis:
I am interested in understanding the factors that influence the rate of crime in cities. And want to compare the data about various cities to understand the variables that affect crime in cities.
Step 1: The first step will be to scrape the data from the above websites
Step 2: Cleaning the data might take significant effort based on how good the data is on the websites.  Predictors such as population, gender ratio, poverty levels need to be created from the raw data that has been scraped 
Step 3: Explore the data to get a feel for which variables might have a correlation with the crime rate
Step 4: Build the linear and polynomial  Regression models to obtain a good prediction model for crime rate for the cities
Step 5: Draw conclusions
References:
Here is a paper about crime in cities which is very informative - http://www.jstor.org/stable/10.1086/250109?seq=1#page_scan_tab_contents
A news article that discuss the same issue - http://www.motherjones.com/kevin-drum/2013/01/are-big-cities-more-dangerous-small-ones/
 https://www.sciencenews.org/blog/science-public/data-driven-crime-prediction-fails-erase-human-bias
https://www.nature.com/news/reform-predictive-policing-1.21338#/b5
KG Muhammads work is very relevant to this topic - http://www.hup.harvard.edu/catalog.php?isbn=9780674062115
Deliverables:
Below are the expected deliverables:
  1. An ipynb notebook showing the work/code
  2. A slide deck summarizing the approach and findings
  3. An algorithm that can be used by city agencies to predict crime in major metropolitan cities in the USA