این سایت در حال حاضر پشتیبانی نمی شود و امکان دارد داده های نشریات بروز نباشند
International Journal of Transportation Engineering، جلد ۱۲، شماره ۲، صفحات ۱۸۱۹-۱۸۴۵
عنوان فارسی
چکیده فارسی مقاله
کلیدواژه‌های فارسی مقاله
عنوان انگلیسی Developing a Framework for Selecting an Appropriate Model based on the Ensemble Learning
چکیده انگلیسی مقاله We present a framework for selecting the optimal ensemble learning model based on 143310 crash observations with five classes. For non-ensemble models, we use five common models. 26 ensemble learning models are derived from these five models. We suggest Diff2 and Diff3 measures for choosing the right model. The diff2 is the difference between observations classified incorrectly as class 1 and incorrectly classified as class 3, 4, or 5. In Diff3, we compare observations misclassified as class 1 or 2 with observations misclassified as class 4 or 5. We select the best model based on the following criteria: for class 1, the largest R1, for class 2, the largest "Diff2", for class 3, a negative "Diff3", and for classes 4 and 5, the highest "F1-score". The paper ranks 31 models based on its criteria. There are five ranking series. By comparing these rankings, we can determine, for example, whether the 3rd best model for class 1 corresponds to the best model for class 2. For each model, 5 "Ranks" are determined. Relationships between the ranks were then evaluated. Rank1 and Rank2, Rank3 and 5 have a relatively strong relationship. A negative and relatively strong correlation exists between Rankings 2 and 3, as well as Rankings 2 and 5.
کلیدواژه‌های انگلیسی مقاله Crash Severity Prediction,Machine learning model,Ensemble Voting Classifier,Imbalanced Multi-Class Classification
نویسندگان مقاله Alireza Mahpour |
Faculty of Civil, Water and Environmental Engineering, Shahid Beheshti University, Tehran, Iran

Mostafa Shafaati |
Faculty of Civil and Environmental Engineering, Tarbiat Modares University, Tehran, Iran

نشانی اینترنتی http://www.ijte.ir/article_207436_70d2fc333661563ccf0001bcc3bb8870.pdf
فایل مقاله فایلی برای مقاله ذخیره نشده است
کد مقاله (doi)
زبان مقاله منتشر شده en
موضوعات مقاله منتشر شده
نوع مقاله منتشر شده
برگشت به: صفحه اول پایگاه   |   نسخه مرتبط   |   نشریه مرتبط   |   فهرست نشریات