As we know, thousands of accidents take place every day. These are very disturbing events which leads to major losses. There are various factors such as weather conditions, non-use of helmets, time period, type of accident, etc. due to which accidents take place. It is necessary to predict the areas which are more accident-prone. This model aims to predict such are considering all the above parameters using Machine Learning Concept i.e. EDA (Exploratory Data Analysis). It can further be improved and we can send the report of the accidents to the authorities like, hospitals, ambulance and insurance agencies and this will be very helpful in bringing down the accident fatality rates in the country.
Road accidents causes a huge impact on the society. As the rate of accidents are increasing with the time, our attention has shifted towards road accidents analysis for determining the factors which significantly affects the accidents.
According to the reports given by World Bank, India accounts for 11% of death due to road accidents globally with only 1% of the world’s vehicles.
 It is important to do analysis of Road accidents to find the causes of accidents and severity of injuries. Many Engineers and researchers have tried to design a model but road accidents are unavoidable. 
Here, we will study and analyse the datasets that are containing details of different states based on various factors such as weather conditions, non-use of helmets, time period, type of accidents, etc. and develop a model that will helps us in preventing road accidents by Exploratory Analysis of Data which will discover the patterns and graphical representations to understand the data better.
The collected data will be analysed and will combine the data on the various parameters using the best algorithm. Through this estimation we can analyse and identify the flaws and reasons of the accidents.
This will be very helpful in making roads so that we can avoid the problems which have been faced earlier. Through this model (Road Accidents Analysis Using Exploratory Data Analysis) we can make roads more secure and accident free.
II. RELATED WORK
S. Shanthi:Introduced a technology based on gender classification i.e. data mining classification, to provide high precision results they used AdaBoost Meta classifier in Rnd Tree and C4.S. The Fatal Analysis Reporting System (FARS) provided The Critical Analysis Reporting Environment (CARE) used by the training data set.
Tessa K. Anderson:Introduced a method of clustering technique that determines stochastic indices which exist in some clusters and can be compared in time and space to identify the high densities accident hotspots. To create the basic spatial unit of the hotspot clustering method, the kernel density estimation tool enables the visualization and manipulation of density based event.
Sachin Kumar:To detect the high frequency accident locations and the root factors that effects road accidents on these locations they used data mining techniques. To identify the relationship between distinct attributes which are in accident data set and to know the characteristics of location the Association Rule mining Algorithm is
Ali Moslah Aljofey:In this paper, they analysed the factors which are responsible for the road accidents on highway locations regularly and tries to Analyse ways to reduce the cause of accidents in such areas. The proposed framework consists of a k-means clustering technique to get the frequencies of highway location accident.
III. PROPOSED METHODOLOGY
In our project we have used Exploratory Data analysis. Exploratory data analysis is an approach of analysing data sets to summarize their main characteristics. It is a critical process of performing initial investigations on data. Used to discover patterns, to spot anomalies, to test hypothesis and to check assumptions with the help of summary statistics and graphical representations.
The exploratory data analysis steps that analysts have in mind when performing EDA include:
Asking the right questions related to the purpose of data analysis
Obtaining in-depth knowledge about problem domains.
Setting clear objectives that are aligned with the desired outcomes.
A. Exploratory Data Analysis
EDA is the process of investigating the dataset to discover patterns, and anomalies (outliers), and form hypotheses based on our understanding of the dataset. It generates graphical representation and summary of statics for numerical data to understand the dataset.
Analysis of Variance is a statistical technique which is used to check if the means of two or more groups are significantly different from each other. It shows the impact of one or more factor by comparing the means of different samples.
The formula for ANOVA is:
Year-wise data set (1970-2017) 2. state-wise data set in 2017 based on: severity of accidents, non-use of helmet, responsibilities of driver, weather condition, time period in a day, type of accidents 
Import all the required libraries such as pandas, Matplotlib, NumPy, etc.
Merge all the data set.
Correlation is check between different feature of Data set.
Perform Exploratory Data Analysis to get the summary and pattern of numerical data in data set.
Apply ANOVA to get the mathematical Value which shows the impact of different Factors
At last, we get the features which are more responsible for Road accident Rate.Then enter the value for these parameters in the form of 0 and 1.
Calculate Percentage of no. of One’s, this percentage is the probability of Accident.
At end we get the probability of Accidents.
a. Uttar Pradesh, Tamil Nadu, Madhya Pradesh, Maharashtra and Karnataka are the top 5 states where most fatal accidents occurred in year 2017.
b. Most of the accidents Occurred in Sunny/Clear weather that too in afternoon and evening time period.
c. And lastly, in Karnataka 4588 people were killed by "hitting from back" and in Tamil Nadu 3507 pedestrian were killed. While, Uttar Pradesh records the most hit and run death cases nearly 3300 in the year 2017.
d. If chances of any 2 features among Midnight, Overnight, exceeding lawful speed, driving on wrong side, it from side is true then probability of Accident is 0.4.
This model of Road Accident Analysis gives us many factors, controlling these factors can leads to decrease in accident rate. The Model uses Exploratory Data Analysis which give pattern in Dataset. There are two different Dataset i.e. state wise and year-wise. When EDA is applied on Year-wise Dataset, it shows that the accident rate increases as per the year. When EDA is applied on State wise Dataset, it shows total accidents in all the state is 1300 average. Tamil Nadu is the state having more accident rates and Uttar Pradesh, Madhya Pradesh, Maharashtra and Karnataka are those states where most fatal accidents occurred in year 2017 due to exceeding lawful speed. Using ANOVA, we have seen that
Midnight, Overnight, exceeding lawful speed, driving on wrong side, hit from side, which have high impact on Fatal Accident rates. So, these features are chosen for finding the Probability of Accidents. User is asked to enter all the 5 parameters value as 1 or 0. If any 2 are true then there is 0.4 probability of Accident. This algorithm gives 95% efficiency and its performance is 1 min. This is the result of this Road Accident Model.
 Chitnis S.D., P Gokhale, Statistical Analysis and Classifier Accuracy Improvements Models for Road Accident Issues on National Highways in India, Design Engineering, Issue 7, ISSN No. 0011-9342, Pages 55765591, 2021.
 Sahil Dabhade, Sai Mahale, Avinash Chitalkar, Pushkar Gawhad, Vicky Pagare, Road Accident Analysis and Prediction using Machine Learning, International Journal for Research in Applied Science & Engineering Technology (IJRASET), ISSN: 23219653; Volume 8 Issue I, Jan 2020
 Ram Prasanth T, Spanglar Diaz V, Surendran N, Udhayavel V, Dr. C. Anand, Mrs. N. Vasuki, A Statistical Analysis of Road Accident Data-Using Data Mining Techniques, ISSN -2278-0181, Volume 6, Issue 08, 2018
 Road Accidents Data of Government of India, 2018 ,https://data.gov.in/catalo g/road- accidents-india2018?filters%5Bfield_catalog_r eference
 ROAD SAFETY FACTS https://www.asirt.org/safetravel/roadcrashes/
 Inderpreet kaur, Ashish Kumar Luhach, Pooja, PROPOSING K-MODE BASEDMETHODOLOGY FOR ROAD ACCIDENTSWITH IMPROVED APRIORI, International Journal of Engineering Applied Sciences and Technology, Vol. 2, Issue 5, ISSN No. 2455-2143, Pages 66-74, 2017
 Cloud analysis data of IBM on Road Accidentshttps://www.ibm.com/cloud/learn/explor at ory-data-analysis
 Gurchetan1000 — A Simple Introduction to ANOVA (with applications in Excel)January 15, 2018 -https://www.analyticsvidhya.com/blog/2018/01/anova-analysis-of-variance/
 Ali Moslah Aljofey, Khalil Alwagih, Analysis of Accident Times for Highway Locations Using K-Means Clustering and Decision Rules Extracted from Decision Trees, International Journal of Computer Applications Technology and Research Volume 7- Issue 01, 001011 , 2018, ISSN:-2319-8656
 Md. Farhan Labib, Ahmed Sady Rifat, Md. Mosabbir Hossain, Amit Kumar Das, Faria Nawrine, Road Accident Analysis and Prediction of Accident Severity by using Machine Learning in Bangladesh, International Conference on Smart computing and communiction, Issue 7, 2019.
 Muthusamy A P, Rajendra M, Ramesh K, Sivaprakash P, A Review on Road Traffic and Related Factors International Journal of Applied Engineering Research, Vol. 10, Issue 7, ISSN No. 0973-4562, Pages 2817728183, 2015.
 Tessa K. Anderson, Kernel density estimation and K-means clustering to profile roadaccident hotspots, Accident Analysis and Prevention Elsevier, 41 (2009) 359–364.
 Shanthi Selvaraj, Geetha Ramani, Classification of seating position specific patterns in road traffic accident data through data mining techniques, Second International Conference on Computer Applications, Pondicherry.
 Sachin Kumar and Durga Toshniwal, A data mining framework to analyse road accident data, Kumar and Toshniwal Journal of Big Data (2015) 2:26 DOI10.1186/s40537-015-0035-y