Hi 👋, I'm Matthew Osadebamwen

A passionate Data Analyst and Data Scientist with crazy interest in Artificial Intelligence

• 🔭 I’m currently working on Analysing Road Traffic Accidents in the UK for the year 2019

• 🌱 I’m also learning Linux, Tableau, Power BI and SQL

• 📫 Email: osadebamwen.matthew@gmail.com

• ⚡ Fun fact: I think I'm good at Chess even though I barely win any game nowadays.

Connect with me:

Languages and Tools:

 

UK Traffic Aaccident Analysis

For this analysis we will use the road safety data available from here: http://data.gov.uk/dataset/road-accidents-safety-data

Introduction

The UK government provides detailed road safety data with respect to injuries, road accidents, type of vehicles involved and casualties. Overall, the data is divided into three datasets: Accidents, Vehicles and Casualties. A summary of each of these datasets is presented in Table 1. The ‘accident index’ is provided in each dataset to identify an accident. I initially proposed to merge all three datasets from the start, before performing the analysis but upon realizing this might be more dimensionally challenging to work with, I decided to work with individual datasets and merge them only when needed. It should also be noted lots of the attributes are categorical hence codes are provided to indicate their respective meanings.

Table 1

                                                                                   

Datasets	Unique Identifier	Number of Attributes	Number of Rows
Accidents	Accident Index	32	117536
Vehicles	Vehicle Reference	23	216381
Casualties	Casualty Reference	16	153158

Aims

It may sound far-fetched to suggest that certain months, days, or hours could be more dangerous. Hence, the aim of this report is to analyse UK accidents data to give insights into the following questions:

(a) Are there significant hours of the day, and days of the week, on which accidents occur?
(b) For motorbikes, are there significant hours of the day, and days of the week, on which accidents occur?
(c) For pedestrians involved in accidents, are there significant hours of the day, and days of the week, on which they are more likely to be involved?
(d) What impact, if any, does daylight savings have on road traffic accidents in the week after it starts and stops?
(e) What impact, if any, does sunrise and sunset times have on road traffic accidents?
(f) Are there particular types of vehicles (engine capacity, age of vehicle, etc.) that are more frequently involved in road traffic accidents?
(g) Are there particular conditions (weather, geographic location, situations) that generate more road traffic accidents?
(h) How does driver related variables affect the outcome (e.g., age of the driver, and the purpose of the journey)?
(i) Can we make predictions about when and where accidents will occur, and the severity of the injuries sustained from the data supplied to improve road safety? How well do our models compare to government models?

Please click on the BigData_&_DataMining.ipynb python note book above to take a look at the detailed analysis.