Dynamic and Interpretable Hazard-Based Models of Traffic Incident Durations

نویسندگان

چکیده

Understanding and predicting the duration or “return-to-normal” time of traffic incidents is important for system-level management optimization road transportation networks. Increasing real-time availability multiple data sources characterizing state urban networks, together with advances in machine learning offer opportunity new improved approaches to this problem that go beyond static statistical analyses incident duration. In paper we consider two such improvements: dynamic update predictions as information about becomes available automated interpretation factors responsible these predictions. For our use case, take one year time-series from M25 motorway London. We it train models predict probability distribution durations, utilizing both time-invariant time-varying features data. The latter allow be updated an progresses, more available. predictions, are fed into Match-Net algorithm, a temporal convolutional hitting-time network, recently developed dynamical survival analysis clinical applications. benchmarked against regression established technique known landmarking found perform favourably by several standard comparison measures. To provide interpretability, utilize concept Shapley values domain interpretable artificial intelligence rank most relevant model at different horizons. example, day always significantly influential feature, whereas strongly influence 5 60-min Although focus here on incidents, methodology describe can applied many problems where combined features.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Incident Duration Modeling Using Flexible Parametric Hazard-Based Models

Assessing and prioritizing the duration time and effects of traffic incidents on major roads present significant challenges for road network managers. This study examines the effect of numerous factors associated with various types of incidents on their duration and proposes an incident duration prediction model. Several parametric accelerated failure time hazard-based models were examined, inc...

متن کامل

Determinants of Inter-trade Durations and Hazard Rates Using Proportional Hazard Arma Models

This paper puts a focus on the hazard function of inter-trade durations to characterize the intraday trading process. It sheds light on the time varying trade intensity and, thus, on the liquidity of an asset and the information channels which propagate price signals among asymmetrically informed market participants. We show, based on an exogenous information process, that the way traders aggre...

متن کامل

Traffic Flow Analysis Based on Queuing Models

One of the most important issues in the plant layout design especially in mass production organizations with high inter-plant logistics is‘material flow and inter-plant traffic analysis and its effects on the production capabilities or pauses in production lines. In this paper the inter-plant traffic analysis issue on the basis of single channel queue model (M/M/1) is analyzed in a carmaker c...

متن کامل

A NOTE TO INTERPRETABLE FUZZY MODELS AND THEIR LEARNING

In this paper we turn the attention to a well developed theory of fuzzy/lin-guis-tic models that are interpretable and, moreover, can be learned from the data.We present four different situations demonstrating both interpretability as well as learning abilities of these models.

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Frontiers in future transportation

سال: 2021

ISSN: ['2673-5210']

DOI: https://doi.org/10.3389/ffutr.2021.669015