Abstract:Scene risk identification is essential for the traffic safety of Internet of Vehicles. However, the performance of existing risk identification approaches is heavily limited by the imbalanced historical data and the poor model interpretability. Meanwhile, the large processing delay and the potential privacy leakage threat also restrict their application. In this paper, a novel risk identification model is proposed that leverages the synthetic minority over-sampling technique nearest neighbor (SMOTEENN) method to balance between high-risk and low-risk data. The risk identification model has fine interpretability by using recursive feature elimination cross validation (RFECV) with the Shapley additive explanation (SHAP) to analyze the importance of different features, and further elaborately design the Focal Loss function to tackle the disparity between the difficult and easy sample learning. The proposed interpretability scene risk identification framework, named iScene, is built on the infrastructure of 6G space-air-ground integrated networks (SAGINs) with blockchain assistance. The model updata efficiency and privacy preservation are effectively enhanced. An elastic computing offloading algorithm is applied to minimize the system overhead under the hierarchical edge service architecture. The experimental evaluation is carried out to verify the effectiveness of the proposed risk identification framework. The results indicate that the G-Mean value is increased by 23.4%, while the task average response delay is reduced by 21.2%, compared to that in the traditional risk identification approaches with local computing services. This article mainly utilizes the LightGBM method, blockchain technology, and an integrated network of space-air-ground to complete the task of risk identification in traffic accident scenarios.image

A Design of Risk Experimental Platform Based the Conception of Screen Scene

Study on Method for Risk Identification and Evaluation in Construction Safety

THE STUDY OF THE CHARACTERISTICS OF ENTREPRENEURIAL RISKY DECISION-MAKING

Iscene: an Interpretable Framework with Hierarchical Edge Services for Scene Risk Identification in 6G Internet of Vehicles

Multi-scenario simulation of landscape ecological risk probability to facilitate different decision-making preferences

From Risk Prediction Models to Risk Assessment Service: A Formulation of Development Paradigm

New Paradigms for the Old Question: Challenging the Expectation Rule Held by Risky Decision-Making Theories

RiskBench: A Scenario-based Benchmark for Risk Identification

Imagery Based Risk Assessment Using Crowdsourcing Technology in Complex Workspaces

Time in hand: Temporal focus in risk discourse and audience emotions on knowledge-sharing platforms

Screen-based 3D Subjective Experiment Software

Stackelberg risk preference design

Rim: Risk Interaction Model And Its Applications

Real-Time Risk Management Based on Time Series Analysis

A Multidisciplinary Review Into the Evolution of Risk Concepts and Their Assessment Methods

Efficient Estimation of a Risk Measure Requiring Two-Stage Simulation Optimization

Regshock: Interactive Visual Analytics of Systemic Risk in Financial Networks

Design and Implementation of an Integrated Risk Assessment System for Chemical Places.

Mapping the fire risk in buildings: A hybrid method of ASET-RSET concept and FED concept.

A Set of Estimation and Decision Preference Experiments for Exploring Risk Assessment Biases in Engineering Students

Risk management in design process for Factory of the future