March 2017

Prepared for review by LinkedIn Analytics

Created by Jerad Acosta

Presentation Keys

Use:

  • -> next slide
  • <- previous slide
  • 'f' to enable full-screen mode
  • 'w' to toggle widescreen mode
  • 'o' to enable overview mode

For context and support please see the companion notebook to see the code and process.

Use the Github repo for a copy of the source code.

Click Here to see a copy of the code used to generate this presentation.

Interact with a custom built dashboard designed for the project stake holder and their intention.

Goal

Reducing Accidents in Brooklyn

The analysis for this deck and the accompanying Data Exploration R Notebook and Dashboard was primarily generated from the NYC Open Data Vehicle Collision Data Set.

Influential factors on this work include:

  • City Council Stake Holders
    • Can leverage their greatest impact in local law and policy (how can law and policy be used to reduce accidents in Brooklyn)
  • Identify factors correlated with accidents
    • Awareness can reduce incident by assisting drivers in pursuing their own interest to keep themselves alive and whole
  • Understand The NYC Collision Data
    • In the act of working with data an analyst can grow an empathy toward its tendencies
      • Maturing their intuition on the matter and adding to a now enhanced skill set for use on this problem as well as future works

The Geography of Events Across NYC

A Visual Exploration…

…A glance at where accidents have been recorded across the five boroughs of New York City…

Hover for more info, Zoom, (de)select Borough Categories…

Events Over Time

Chart Accident Count for each hour of each day of the week in order to:
  • Visually search for patterns
    • Take advantage of priori knowledge (e.g. human nature toward a weekly schedule & routine)
    • People live, work and plan around a 7-day weekly schedule that has universal social influencers
    • For example: a 9-5 work-schedule or Weekends

Hours of the Week

Daily patterns across the week and weekly patterns within each day

Zoom, Rotate, Hover, Pan, …

Exploring Weather

Weather Data was collected from the NOAA API.
Data on snow and rain fall was merged with accident data to search for what seems like an intuitive correlation.
Oddly, no significant correlation was found; However, upon inspecting the data seasonally as was done for the weather, a pattern becomes clear. The pattern is across time and appears to be annual, consistently growing from 2012 to 2015 in all seasons and through to 2016 in Spring and Winter.
Thus, we can support the premise of exploring extreme options with regard to vehicle safety in Brooklyn - including legal and campaigning.

Continue The Journey

Further Your Exploration

Checkout the code to this deck, the exploratory work and tutorial used to create it, and an interactive dashboard to further explore the data set



Click my profile image to see more of my projects on my website: Jerad.xyz




Connect with me on LinkedIn









Take it for a spin!

An interactive dashboard custom designed, coded and hosted to support our client and stakeholder's goal of decreasing car accidents in Brooklyn.

This dashboard will be accessable to the client anywhere they have internet access so they can recreate the graphics here as well as conduct their own queries in support of their goal.

Supporting Documents