| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

Show document

Title:Modeliranje prostorskega vzorca vremensko pogojenih cestnoprometnih nesreč v Sloveniji
Authors:Horvat, Nina (Author)
Ivajnšič, Danijel (Mentor) More about this mentor... New window
Žiberna, Igor (Co-mentor)
Files:.pdf MAG_Horvat_Nina_2019.pdf (6,11 MB)
MD5: F22700B963597B806CE3DAD506FA143B
 
Language:Slovenian
Work type:Master's thesis/paper (mb22)
Typology:2.09 - Master's Thesis
Organization:FF - Faculty of Arts
Abstract:Namen magistrskega dela je bil preučiti trende in prostorski vzorec prometnih nesreč, ki so se v obdobju 2006–2017 končale s smrtjo ali hudo telesno poškodbo, v vseh slovenskih občinah in po posameznih vremenskih situacijah (dež, megla, veter, sneg, toča, jasno in oblačno vreme). Naši cilji so bili predstaviti prometne nesreče na splošno, pregledati trende prometnih nesreč na svetovni, evropski in državni ravni, predstaviti vpliv vremena na prometne nesreče, preučiti metode za proučevanje prostorskih vzorcev prometnih nesreč ter analizirati prostorski vzorec izbranih prometnih nesreč. Podatke o prometnih nesrečah smo pridobili na spletni strani Javne agencije Republike Slovenije za varnost prometa in na Statističnem uradu Republike Slovenije, obdelovali smo jih s pomočjo programov Excel, ArcGIS 10.5 in R. Za vsako omenjeno vremensko situacijo smo izdelali prostorski vzorec, trend in standardizirano stopnjo prometnih nesreč. Izbrali smo spremenljivke, ki značilno vplivajo na število prometnih nesreč. Le-te smo v nadaljevanju transformirali s faktorsko analizo, da smo zagotovili pogoje uporabe linearnih modelov (nekoreliranost prediktorskih spremenljivk). Odvisne in neodvisne spremenljivke smo vstavili v model GWR, za katerega pa se je izkazalo, da ni dober napovedovalec števila prometnih nesreč, ko gre za posamezno vremensko situacijo. Posledično smo uporabili še metodo gradnje odločitvenega drevesa (ang. decision tree model). Sestavili smo algoritme, ki z zadovoljivo točnostjo napovedujejo število prometnih nesreč v posamezni občini glede na obravnavano vremensko situacijo. Kakovost modelov odločitvenega drevesa smo izmerili s pomočjo kazalcev: pojasnjen odklon (ang. explained deviance), MAE (ang. mean absolute error), RMSE (ang. ruth mean square error) in Moranov indeks avtokorelacije (ang. Moran's Index). Rezultati naloge imajo aplikativen pomen za snovalce strategije prometne varnosti na nacionalnem nivoju.
Keywords:prometne nesreče, prostorski vzorec, vremenske okoliščine, geografsko obtežena regresija, odločitveno drevo
Year of publishing:2019
Publisher:[N. Horvat]
Source:Maribor
UDC:614.86:551.515(497.4)(043.2)
COBISS_ID:24697096 New window
NUK URN:URN:SI:UM:DK:BJQ48YQ5
Views:458
Downloads:142
Metadata:XML RDF-CHPDL DC-XML DC-RDF
Categories:FF
:
  
Average score:(0 votes)
Your score:Voting is allowed only for logged in users.
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Link:http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:06.06.2019

Secondary language

Language:English
Title:Modelling the Geospatial Pattern of Weather-Related Road Traffic Accidents in Slovenia
Abstract:The purpose of the master’s thesis was to study trends and spatial pattern of road accidents in period from 2006 to 2017, which ended in death or serious physical injuries, in all slovenian municipalities, in every weather condition (rain, fog, wind, snow, hail, clear and cloudy weather). Our purpose was to present road accidents in general, to review trends of road accident at international, at european and national level; furthermore, to present the impact of the weather on road accidents, to examine the methods of spatial patterns of traffic accidents and to analyse the spatial pattern of chosen accidents. Traffic accidents data were obtained at the website of Slovenian Traffic Safety Agency and Statistical Office of the Republic of Slovenia. We analysed the data with different programs, including Excel, ArcGIS 10.5 and R. For every weather condition given we made a spatial pattern, trend and standardized level of traffic accidents. We have chosen variables with significant impact on the number of road accidents. Furthermore we transformed those variables with factor analysis in order to provide conditions for the use of linear models (uncorrelated predictor variables). We inserted dependent and independent variables into the GWR model, which later resulted in not being a good predictor of the number of traffic accidents based on individual weather situation. Consequently, we used the method of building a decision tree model. We have compiled algorithms that predict the accuracy of traffic accidents, in a particular municipality, according to the specific weather situation. As a result, we measured the quality of the decision tree models using indicators: explanatory theories of deviance, MAE (mean absolute error), RMSE (root mean square error ) and Moran's Index. To conclude, the results of the task have an applicative significance for the designers of the road safety strategy at the national level.
Keywords:Road Accidents, Spatial Pattern, Weather Conditions, Geographically Weighted Regression, Decision Tree Model


Comments

Leave comment

You have to log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back
Logos of partners University of Maribor University of Ljubljana University of Primorska University of Nova Gorica