ANALYSIS OF ROAD ACCIDENTS USING DATA VISUALIZATION TECHNIQUES

DATA VISUALIZATION (CSE3020)

J COMPONENT
REVIEW REPORT

TEAM MEMBERS

Student Name Registration Number

Y D S S Prathyusha 18BCE2051
Siddhant Chourasia 18BCE0275
Vaibhav Singh 18BCE2313

Submitted to

Prof. Meenakshi S.P

1
 LIST OF TOPICS

Chapter No Topic Page No

1 INTRODUCTION

1.1 Abstract 3

1.2 Introduction 3

1.3 Objectives 3

2 PROBLEM STATEMENT 3

3 DATA COLLECTION

3.1 Dataset view 4

3.2 Description 5

4 TASKS, SYSTEM DESIGN 5

& TOOLS USED

5 ACTIONS 7

6 IMPLEMENTATION 7

7 VALIDATION

7.1 Domain level 22

7.2 Task level 22

7.3 Data level 22

7.4 Algorithm level 22

8 SUGGESTIONS/ 23
CONCLUSION

2
1.INTRODUCTION

1.1 ABSTRACT

Roadway traffic safety is a major concern for transportation governing agencies as well as
ordinary citizens. In order to give safe driving suggestions, careful analysis of roadway traffic
data is critical to find out variables that are closely related to fatal accidents. The relationship
between fatal rate and other attributes including collision manner, weather, surface condition,
causality severity, and driver age were investigated.
Association rules were discovered by Apriori algorithm, classification model was built by Naive
Bayes classifier, and clusters were formed by a simple K-means clustering algorithm. Certain
safety driving suggestions were made based on statistics, association rules, classification model,
and clusters obtained.

1.2 INTRODUCTION

Investigations of the high-risk areas for road traffic crashes (RTCs) are urgently needed to guide
improvements in road safety. In this paper we apply statistics analysis and data visualisation
algorithms on the FARS Fatal Accident dataset as an attempt to address this problem. In this
paper, the relationship between fatality rate and other attributes including collision manner,
weather, surface condition, causality severity, and driver age were investigated.

1.3 OBJECTIVE

● Our Objective is to arrive at a conclusion regarding the problem of Roadway traffic

safety and to give safe driving suggestions.
● .Finding Numbers of accidents happening in rural and urban areas.
● Finding the trend in the number of accidents that occur each year.
● .Finding the most dangerous time to drive.
● Finding What fraction of accidents caused minor injuries , major injuries and deaths.
● Finding how fast the number of car accidents drop off with age.
● Finding the ratio of men and women who get injured in accidents.

2. PROBLEM STATEMENT

Roadway traffic safety is a major concern for transportation governing agencies as well as
ordinary citizens. In order to give safe driving suggestions, careful analysis of roadway traffic
data is critical to find out variables that are closely related to fatal accidents. India is a country

3
having a high usage of vehicles. The vehicle consumption has drastically increased in the last 40
years from 6 million to 230 million vehicles. Due to the increasing rate of 9% vehicles per year,
the occurrence of road accidents has increased exponentially which in turn has hampered the
road security of the people in India.

3. DATA COLLECTION

3.1 DATASET VIEW

4
3.2 DESCRIPTION

Categorical attributes present – State/Union Territory

Quantitative attributes present- All Other attributes present based on each database

For example

Quantitative attributes present in

“Details_of_road_accident_deaths_by_situation_state_2014.csv”

Year, CrimeHeadOffenders (Driver/Pedestrian), Died_MaleOffenders (Driver/Pedestrian),

Died_FemaleOffenders (Driver/Pedestrian), Died_TransgenderOffenders (Driver/Pedestrian) ,
Died_TotalVictims, Died_MaleVictims, Died_FemaleVictims, Died_Transgender, Victims
Died_Total

4. TASKS

● Download datasets
● Data preparation
● Modelling

5
 ● Visualization

SYSTEM DESIGN

TOOLS/ IMPORTANT LIBRARIES USED

Numpy -NumPy is a Python library used for working with arrays. It also has functions for
working in the domain of linear algebra, fourier transform, and matrices.

Pandas - Pandas is a software library written for the Python programming language for data
manipulation and analysis. In particular, it offers data structures and operations for manipulating
numerical tables and time series.

Matplotlib- Matplotlib is a plotting library for the Python programming language and its
numerical mathematics extension NumPy. It provides an object-oriented API for embedding
plots into applications.

Seaborn- Seaborn is a Python data visualization library based on matplotlib. It provides a

high-level interface for drawing attractive and informative statistical graphics.

Plotly- The plotly Python library is an interactive, open-source plotting library that supports over
40 unique chart types covering a wide range of statistical, financial, geographic, scientific, and
3-dimensional use-cases.

6
OrderedDict- OrderedDict is a dict subclass that preserves the order in which key-value pairs,
commonly known as items, are inserted into the dictionary.

Datetime- It is a Python library used for Time access and conversions.

Bar_chart_race- Make animated bar chart races in Python with matplotlib. Img.

5. ACTIONS

● Data pre-processing
● Feature selection
● Training the model
● Statistical analysis
● Testing
● Output/ graphs prediction

6. IMPLEMENTATION

7
Percentage of accidents per 3-hour period

8
Number of Accidents happening in particular time interval

9
 Number of Accidents at various severity level with respect to speed zone

Number of offenders died in accidents with respect to gender

10
Number of victims died with respect to gender

11
12
Percentage change in the accidents with previous year

13
14
15
16
17
18
19
20
21
7. VALIDATION

7.1 DOMAIN LEVEL

● Our primary aim is to analyse the data obtained by applying statistics analysis and data
visualisation algorithms on various Accident datasets.
● Certain safety driving suggestions were made based on statistics, association rules,
classification model, and clusters obtained.
● These are made by analysing data and comparing trends to find out why and how such
factors can affect road accidents.

7.2 TASK LEVEL

● The main motivation of our project revolves around the fact that in India approximately
465033 people lose their lives every year because of road accidents.
● India’s young, productive population, aged 18-45 years, is involved in 70% of road
accidents.
● So, through this project we aim to analyse trends and information /data related to road
accidents in India and give a detailed analysis of these accidents which may be very
helpful in controlling these accidents.
● This project analysis makes people cautious of the way accidents happen so that they can
take precautions while driving.

7.3 DATA LEVEL

● Reasons for accidents.

● Percentage of accidents in particular periods.
● Types of vehicles involved.
● Severity index.
● Reason for accidents.
● Road accidents of all states and union territories from 2001-14.
● Road accidents on a yearly basis.
● Monthly analysis on states having the highest number of accidents.
● Analysis based on drivers age.

7.4 ALGORITHM LEVEL

● Language used : python

● IDE used - spyder and google collab

22
8. SUGGESTIONS/ CONCLUSIONS

● Whoever is driving should make sure that he is in a proper state to drive.

● Most of the accidents were happening in the time interval of 3pm-6pm, so people should
be careful for various reasons like change in light condition etc..
● Apart from the year 2008, in all other years there is either no change or increase in the
percentage of accidents . Despite so many warnings and laws issued regarding Road
accidents like no triple riding, don’t drink and drive, carry the prescribed weight on
vehicles, everybody finds a way to violate these rules.But people should understand that
at the end it’s going to risk their lives and the lives of people travelling with them.
● The highest number of accidents are recorded from age groups 15-24 and 25-64. There
can be various reasons for this like : Alcoholism, Bike racing, Reckless driving, or the
urge to work or college on time etc...
● We have also seen that in the months of the rainy season the accidents actually peak
down because people are extra careful.
● How many guidelines or awareness drives the government or any organization start , they
are of no use until the people consider them seriously. Value their lives and other’s lives.
● Even if people are following every precaution there is always a loophole in problems like
this. So all we can do is be careful, be sober.

23