Urban Anomaly Analytics: Description, Detection, and Prediction

Abstract

Urban anomalies may result in loss of life or property if not handled properly. Automatically alerting anomalies in their early stage or even predicting anomalies before happening are of great value for populations. Recently, data-driven urban anomaly analysis frameworks have been forming, which utilize urban big data and machine learning algorithms to detect and predict urban anomalies automatically. In this survey, we make a comprehensive review of the state-of-the-art research on urban anomaly analytics. We first give an overview of four main types of urban anomalies, traffic anomaly, unexpected crowds, environment anomaly, and individual anomaly. Next, we summarize various types of urban datasets obtained from diverse devices, i.e., trajectory, trip records, CDRs, urban sensors, event records, environment data, social media and surveillance cameras. Subsequently, a comprehensive survey of issues on detecting and predicting techniques for urban anomalies is presented. Finally, research challenges and open problems as discussed.

Low-Rank on Graphs Plus Temporally Smooth Sparse Decomposition for Anomaly Detection in Spatiotemporal Data

“…The events that correspond to the selected tensor elements are classified as detected. In previous work, similar case studies were presented for experiments on real data [29,30,31,27]. The results are reported in Table 2.…”

Graph Neural Networks for Anomaly Detection in Industrial Internet of Things

“…3, the traffic anomalies include abnormal traffic congestion, traffic accident, damaged traffic infrastructure (e.g., a road). Traffic anomaly detection plays a critical role in fostering the urban ITS [82]. In particular, traffic anomalies are often indicators of traffic accidents, traffic congestion, traffic violations, and damaged traffic infrastructure [83].…”

Recent Advances in Wearable Sensing Technologies

“…When wearables are used in remote services collecting data from users to estimate environmental variables of interest (e.g., pollution, noise, or road traffic levels) through participatory or crowdsensing systems, spoofing attacks can be mitigated at the remote service by taking advantage of the redundant data collected by multiple users to estimate and filter out incorrect, erroneous, or fake data. Methods to filter out data in remote services include kriging, principal component analysis (PCA), Markov random fields (MRFs), Gaussian mixture models (GMMs), stochastic processes [154][155][156][157][158], and anomaly detection algorithms based on ML methods such as support vector machines (SVMs), neural networks (NNs) [159], and recent methods based on DL (i.e., convolutional neural networks (CNNs), and long short-term memory (LSTM) neural networks [160][161][162]).…”