Correcting and complementing freeway traffic accident data using Mahalanobis distance based outlier detection
| العنوان: | Correcting and complementing freeway traffic accident data using Mahalanobis distance based outlier detection |
|---|---|
| المؤلفون: | Guohua Bai, Prashant Goswami, Bin Sun, Wei Cheng |
| المصدر: | Tehnički vjesnik Volume 24 Issue 5 |
| بيانات النشر: | Mechanical Engineering Faculty in Slavonski Brod, 2017. |
| سنة النشر: | 2017 |
| مصطلحات موضوعية: | 050210 logistics & transportation, Mahalanobis distance, Computer science, business.industry, Traffic accident, 05 social sciences, Supervised learning, General Engineering, accident data, data labelling, differential distance, outlier detection, traffic data, updatable algorithm, algoritam kojeg je moguće ažurirati, diferencijalna udaljenost, Mahalanobis udaljenost, otkrivanje outliera, označavanje podataka, podaci o nesreći, podaci o prometu, ComputerApplications_COMPUTERSINOTHERSYSTEMS, 02 engineering and technology, Machine learning, computer.software_genre, Accident (fallacy), ComputingMethodologies_PATTERNRECOGNITION, 0502 economics and business, 0202 electrical engineering, electronic engineering, information engineering, 020201 artificial intelligence & image processing, Anomaly detection, Artificial intelligence, Data mining, business, computer |
| الوصف: | Arhivirana je ogromna količina podataka o prometu koji bi se mogli koristiti za dobivanje specifičnih podataka. Međutim, oni se u potpunosti ne koriste zbog nepostojanja točnih podataka o prometu (oznaka). U ovom radu poboljšavamo algoritam zasnovan na Mahalanobis udaljenosti za procjenu promjena toka prometa i otkrivanje nesreća i primjenjujemo ga kod ispravljanja i dopunjavanja informacija o nesreći. Algoritam za otkrivanje outliera (netipičnih vrijednosti) pruža točne podatke o vremenu događanja nesreće, trajanju i smjeru. Razvijamo i sustav s interaktivnim sučeljem korisnika u svrhu ostvarenja ovog postupka. Predlažu se tri načina za manipulaciju podacima. Najprije, za otkrivanje outliera u prometu predlažemo uporabu multi-metričkih podataka o prometu umjesto jedno metričkih. Nadalje, predlažemo praktičnu metodu za organizaciju prometnih podataka i evaluaciju organizacije Mahalanobis udaljenosti. Kao treće, dajemo opis opće metode za modifikaciju algoritama Mahalanobis udaljenosti kako bi se mogli ažurirati. A huge amount of traffic data is archived which can be used in data mining especially supervised learning. However, it is not being fully used due to lack of accurate accident information (labels). In this study, we improve a Mahalanobis distance based algorithm to be able to handle differential data to estimate flow fluctuations and detect accidents and use it to support correcting and complementing accident information. The outlier detection algorithm provides accurate suggestions for accident occurring time, duration and direction. We also develop a system with interactive user interface to realize this procedure. There are three contributions for data handling. Firstly, we propose to use multi-metric traffic data instead of single metric for traffic outlier detection. Secondly, we present a practical method to organise traffic data and to evaluate the organisation for Mahalanobis distance. Thirdly, we describe a general method to modify Mahalanobis distance algorithms to be updatable. |
| وصف الملف: | application/pdf |
| تدمد: | 1848-6339 1330-3651 |
| DOI: | 10.17559/tv-20150616163905 |
| URL الوصول: | https://explore.openaire.eu/search/publication?articleId=doi_dedup___::68254a41d987e277b42f51ca4ed4a3d7 https://doi.org/10.17559/tv-20150616163905 |
| Rights: | OPEN |
| رقم الانضمام: | edsair.doi.dedup.....68254a41d987e277b42f51ca4ed4a3d7 |
| قاعدة البيانات: | OpenAIRE |
| تدمد: | 18486339 13303651 |
|---|---|
| DOI: | 10.17559/tv-20150616163905 |