Regression Analysis Of The Rate Motor Vehicle Accidents In India

Abstract

Motor vehicle accident in unexpected event which produce intended injury death, produce damage to human lives involving one or more vehicle on road the. It is explicitly indisputable that motor vehicle accidents have increasingly become a major cause of concern for highway safety engineers and transportation agencies in

Big cities over the last few decades. The focus of this study is to analyse the rate of report motor vehicle accidents in big cities as considering the period in which these accidents occur. The analyse of data is carried using of percentage, regression and ANOVA in R software the data now in a monthly return submitted by the driver and the vehicle licensing department. The analysis will help to get insight on the various factor leading for motor accidents and would help to further decision according related to traffic policy.

Keywords: Regression, r software, r algorithm

Indtroduction

Deaths resulting from road accidents have become a big problem in the under developing countries. Accident is an unplanned and unexpected event usually resulting in injury or property damage. Vehicular accidents in big cities have become one of the worrying and growing concerns in recent times. This is as a result of the unbelievable negative effects of road traffic accidents on human lives, properties and the environment. We are focusing on road traffic accidents and have to find out the main problem behind this accident. Road accident are very critical not only because they result in harm and ensuing disability or death of people, but also because they result in waste of resources such as those hospital services that could be used for other purposes, and loss of savings and working days of the accident victims which may improve the future life style of them families. The most dominant factor in understanding the chain of events leading to an accident is the human factor and driver’s perceptions of the issue are necessary for interpositions to be effective. Some countries have provided creative methods of educating people. Internet “talk – back” to accident news are used as an innovative method that supplement qualitative techniques such as focus group or interviews. Motor-vehicle is a single greatest cause of accidents in the developed world. The huge rise in the number of motor cars in the 20th century has resulted in approximately 5,000 deaths a year from vehicle accident.

One serious road accident in the country occurs every minute and 16 die on Indian roads every hour. 1214 road crashes occur every day in India.

Two wheelers account for 25% of total road crash deaths.20 children under the age of 14 die every day due to road crashes in in the country. 377 people die every day, equivalent to a jumbo jet crashing every day. Two people die every hour in Uttar Pradesh – State with maximum number of road crash deaths. Tamil Nadu is the state with the maximum number of road crash injuries

Top 10 Cities with the highest number of Road Crash Deaths (Rank –Wise):

1. Delhi (City)
2. Chennai
3. Jaipur
4. Bengaluru
5. Mumbai
6. Kanpur
7. Lucknow
8. Agra
9. Hyderabad
10. Pune

Literature Survey

Road accidents in the country have decreased by around 4.1% during 2016, with the year seeing 4,80,652 road accidents as against 5,01,423 in 2015. However, fatalities resulting from these accidents have risen by about 3.2% during the same period. Nearly 1,50,785 persons were killed in 2016 as against 1,46,133 in 2015. Releasing the Ministry’s annual publication ‘Road Accidents in India-2016. Road Accidents in India, 2016 is a compilation of data on various facets of road accidents as furnished by Police Departments of all States/U.Ts for the calendar year 2016. Shri Gadkari also informed that the positive trend of 2016 is further bolstered by the accident figures for the first half of 2017, where there has been a 3 % reduction in road accidents between January to July 2017, along with a 4.75 % reduction in road accident fatalities. While road accidents have come down from 2,43,870 between January to July 2016 to 2,36,458 during the same period in 2017, fatalities have come down from 79,354 between January to July 2016 to 75,583 during the same period in 2017. Road accidents deaths have reduced in 25 states and Union Territories in the first half of 2017. Only states like Assam, Bihar, Orissa and Uttar Pradesh seen increase in road accident fatalities between 2-8 % during this period. The National Highways accounted for 29.6 per cent of total road accidents and 34.5 per cent of total number of persons killed. As compared to the previous year i.e, 2015 road accident has gone up on National highways from 28.4 per cent to 29.6 per cent in 2016. The State Highways accounted for 25.3 per cent of total accidents and 27.9 per cent of the total number of persons killed in road accident in 2016.

Road Accidents - a leading cause of Injuries, deaths & disabilities:

India : 2016

Accidents – 4,80,652

Deaths – 1,50,785

Person Injured – 4,94,624

In Every Day: 1,317 Accidents take place and 413 Persons killed on Indian Roads

In Every Hour: 55 Accidents take place and 17 Persons killed on Indian Roads

Methodology

A simple linear regression equation of the dependent variable on each of the other factors and a multiple regression equation was fitted on all the independent variables. We use linear regression as well as ANOVA. For finding number of road accident in per year. By using R software.

Save your time!
We can take care of your essay

• Proper editing and formatting
• Free revision, title page, and bibliography
• Flexible prices and money-back guarantee

Place Order

Hypothesis of the study

Null Hypothesis (Ho): There is no significant relationship between road traffic accidents and population growth in big cities.

Alternative Hypothesis (H1): There is a significant relationship between road traffic accidents and population growth in big cities.

Lasted five year data 2011-2016 implemented R software

Apriori Algorithm:

Apriori Association rule algorithm is used for mining the frequent patterns in database where support and confidence are the two measures used to measure association rule quality . Support is the percentage of transaction in the database that contains XUY for the association rule X->Y . Confidence is the ratio of the number of transaction that contains XUY to the number of transaction that contains

Predictive Apriori Algorithm:

Predictive APRIORI association rule algorithm is also used for mining patterns in database. It differs than Apriori algorithm in that both support and confidence measures are combined into a single measure called predictive accuracy.

Experimental Setup:

Our research is based on the Real-life traffic accidents records collected from Indian Traffic Department from 2008 to 2010.

Data Preprocessing:

After collecting traffic accident records from the traffic department in Indian Police Authority, data were cleaned and preprocessed. The problem of having many missing values in many records was solved by ignoring all those records with missing or incomplete values.

Selection of Attributes and Records:

After pre-processing, 17 attributes covering three types of factors (Accident, Driver, and Road) were developed. Our focus on this research was on three types of severity classes of accidents; death, severe, and moderate.

Experimental Results

There are four subsections to focus on when discussing the results related to the analysis; the generated rules discovered from both association rules algorithms for each of the three classes with summary of these rules and the comparison of the rules extracted for each class from each algorithm to find out which algorithm is more suitable for traffic accident analysis. The rules for the years 2008 to 2010 are generated for each of the three classes; death, severe, and moderate using Apriori and Predictive Apriori algorithms. Best twenty rules were first generated for death class, severe class, and moderate class by WEKA. The summarized final sets of rules for each class were then generated using rule covers method. A rule covers method is applied to summarize the best rules generated by Apriori algorithm. It summarizes the rules by removing shorter rules covered in bigger rules. For the rules generated by Predictive Apriori algorithm, we combined the rules and calculated the accuracy for the each summarized rule.

Discussions And Recommendations

Based on this research, we can observe that when applying rule covers method on the best rules for the three classes generated (death, severe, and moderate) using Apriori algorithm, few summarized rules were obtained after eliminating shorter rules covered in longer ones. On the other hand, when we applied rule covers method on the best rules for the three classes using Predictive Apriori algorithm, no rules were eliminated. We could only aggregate them together. In addition, the best twenty rules generated by Apriori algorithm contained combinations of various accidents’ factors unlike the best twenty rules generated by Predictive Apriori algorithm where each rule contained a single factor at a time and lacked the associations among the different accident factors. Therefore, empirical results showed that class association rules generated by Apriori algorithm were more effective than those generated by Predictive Apriori algorithm if associations between the different factors are of high significance. More number of rules could be eliminated and more associations between accident factors and accident severity were explored when applying Apriori algorithm.

Conclusion

Our experiments showed that when applying rule covers method on the generated class association rules using Apriori and Predictive Apriori algorithms, many class association rules generated by Apriori algorithm were eliminated and more effective rules than those generated by Predictive Apriori algorithm were obtained. In addition, more associations between accident factors and accident severity were explored when applying Apriori algorithm. On the other hand, Predictive Apriori algorithm could derive more number of rules that could be useful when studying the effect of each individual factor to accident severity. The adaptation of the association rule mining algorithm to mine only a particular subset of association rules where the classification class attribute is assigned to the right-hand side could successfully generate more effective rules covering all three classes.using 5 year of accidental data predictive by use of r software .

References

1. https://www.researchgate.net/publication/257717424_Using_Decision_Trees_to_Extract_Decision_Rules_from_Police_Reports_on_Road_Accidents/download
2. IISTE journals can find the submission instruction on the following page: http://www.iiste.org/Journals/
3. P Sprent. Models in regression and related topics. Methuen's monographson applied probability and statistics. Methuen, London 1969. 45”
4. Bhatthcharyya G. K and Johnson R.A (1977) Statistical Concept and Methods pp. 428-444 John Wiley andSons, New York.