Deep Temporal Multi-Graph Convolutional Network for Crime Prediction

Wang, Yaqian; Ge, Liangpeng; Li, Siyu; Chang, Feng

doi:10.1007/978-3-030-62522-1_39

Cited by 16 publications

(17 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Temporal Convolutional Networks (TCN) is a special convolutional neural network, which successfully implements the application of a convolutional neural network in time series problems. The convolution kernel size limits the performance of the traditional convolutional neural network in time series problems, leading to its inability to fully extract sequence features [25]. Therefore, when dealing with time series problems, the recurrent neural network (RNN) is usually used for modeling.…”

Section: Tcnmentioning

confidence: 99%

A New Ensemble Reinforcement Learning Recursive Network for Traffic Volume Forecasting in a Freeway Network

Yan

Dong

et al. 2022

Advances in Transdisciplinary Engineering

View full text Add to dashboard Cite

While recent development, the freeways promote economic and traffic demand, but it also increases a lot of traffic congestion and accidents. The effective traffic volume forecasting technology can reduce traffic congestion, and improve traffic network planning and information management. In this research, a new ensemble forecasting model is proposed to forecast traffic volume. The framework of the model consists of the TCN model, the GRU model, and the ensemble SARSA algorithm. TCN and GRU are utilized as predictors to construct the prediction model for deep mining of the traffic volume data and SARSA can integrate the prediction results from these two predictors and achieve satisfactory results. We used two sets of the traffic data collected from a freeway in Changsha to verify the performance. The experiment results proved the effectiveness of the hybrid model that had the best accuracy in both cases, and the MAPEs of TCN-GRU-Sarsa are 7.8281% and 6.9015%.

show abstract

Section: Tcnmentioning

confidence: 99%

A New Ensemble Reinforcement Learning Recursive Network for Traffic Volume Forecasting in a Freeway Network

Yan

Dong

et al. 2022

Advances in Transdisciplinary Engineering

View full text Add to dashboard Cite

show abstract

“…We introduce some external data sources that have been shown to be helpful in crime prediction. Environmental context is often considered in crime predictive studies, such as meteorological 2 https://data.cityofnewyork.us/Public-Safety/NYC-crime/qb7u-rbmr 3 https://data.cityofchicago.org/Public-Safety/Crimes-2019/w98m-zvie 4 https://data.sfgov.org/Public-Safety/Police-Department-Incident-Reports-2018-to-Present/wg3w-h783 5 https://www.atlantapd.org/i-want-to/crime-data-downloads 6 https://www.opendataphilly.org/dataset/crime-incidents 7 https://data.baltimorecity.gov/datasets/part1-crime-data/explore 8 https://geodash.vpd.ca/opendata/ 9 https://data.london.gov.uk/dataset/mps-homicide-dashboard-data 10 https://www.kaggle.com/inquisitivecrow/crime-data-in-brazil 11 https://data.london.gov.uk/dataset/mps-homicide-dashboard-data 12 https://data.london.gov.uk/dataset/mps-business-crime-dashboard-data 13 data (e.g., temperature and weather) [127,136,139], demographic data (e.g., median age and race ratio) [14,15,134], geographic data (e.g., longitude and latitude) [134], Point-of-Interests (POI) data (e.g., shopping, sports and education) [14,55,136], urban environmental data (e.g., noise, traffic flow, taxi trip) [14,103,136,139], and human behavior data (i.e., mobile data) [15,103,156]. A POI is a record of a place on a map that someone finds useful or interesting, typically defined by its geographical coordinates and a few additional attributes like name and category.…”

Section: Open Source Indicatorsmentioning

confidence: 99%

“…Other studies have applied GNN models to knowledge graph-based data to enhance contextual information for event prediction by learning multi-relational features [35,36]. For crime prediction, some work employed GNN models to enhance representation learning via capturing spatial correlations among pre-defined graph nodes, thereby improving forecasting accuracy [62,136,139,143].…”

Section: Deep Learning Techniquesmentioning

confidence: 99%

“…• Leveraging Heterogeneous Data. As the availability of various open-source data continues to increase, researchers have begun to resort to heterogeneous data to develop predictive models [86,119,136]. Societal events occur in a dynamic social environment, and their corresponding key information can publish in various forms, such as traditional news reports, social media, and government official reports.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Survey on Societal Event Forecasting with Deep Learning

Deng¹,

Ning²

2021

Preprint

View full text Add to dashboard Cite

Population-level societal events, such as civil unrest and crime, often have a significant impact on our daily life. Forecasting such events is of great importance for decision-making and resource allocation. Event prediction has traditionally been challenging due to the lack of knowledge regarding the true causes and underlying mechanisms of event occurrence. In recent years, research on event forecasting has made significant progress due to two main reasons: (1) the development of machine learning and deep learning algorithms and (2) the accessibility of public data such as social media, news sources, blogs, economic indicators, and other meta-data sources. The explosive growth of data and the remarkable advancement in software/hardware technologies have led to applications of deep learning techniques in societal event studies. This paper is dedicated to providing a systematic and comprehensive overview of deep learning technologies for societal event predictions. We focus on two domains of societal events: civil unrest and crime. We first introduce how event forecasting problems are formulated as a machine learning prediction task. Then, we summarize data resources, traditional methods, and recent development of deep learning models for these problems. Finally, we discuss the challenges in societal event forecasting and put forward some promising directions for future research.

show abstract

“…Graphs are widely used in various applications, such as crime forecasting system [18,38], recommendation systems [42,43], social networks, and spam detection [15,26]. For those non-Euclidean data, Graph Neural Networks (GNNs) are powerful architectures for graph representation and learning.…”

Section: Introductionmentioning

confidence: 99%

Subgroup Fairness in Graph-based Spam Detection

Liu¹,

Lyu²,

Zhang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Fake reviews are prevalent on review websites such as Amazon and Yelp. GNN is the state-of-the-art method that can detect suspicious reviewers by exploiting the topologies of the graph connecting reviewers, reviews, and target products. However, the discrepancy in the detection accuracy over different groups of reviewers causes discriminative treatment of different reviewers of the websites, leading to less engagement and trustworthiness of such websites. The complex dependencies over the review graph introduce difficulties in teasing out subgroups of reviewers that are hidden within larger groups and are treated unfairly. There is no previous study that defines and discovers the subtle subgroups to improve equitable treatment of reviewers. In this paper, we address the challenges of defining, discovering, and utilizing subgroup memberships for fair spam detection. We first define a subgroup membership that can lead to discrepant accuracy in the subgroups. Since the subgroup membership is usually not observable while also important to guide the GNN detector to balance the treatment, we design a model that jointly infers the hidden subgroup memberships and exploits the membership for calibrating the target GNN's detection accuracy across subgroups. Comprehensive results on two large Yelp review datasets demonstrate that the proposed model can be trained to treat the subgroups more fairly.

show abstract

Deep Temporal Multi-Graph Convolutional Network for Crime Prediction

Cited by 16 publications

References 13 publications

A New Ensemble Reinforcement Learning Recursive Network for Traffic Volume Forecasting in a Freeway Network

A New Ensemble Reinforcement Learning Recursive Network for Traffic Volume Forecasting in a Freeway Network

A Survey on Societal Event Forecasting with Deep Learning

Subgroup Fairness in Graph-based Spam Detection

Contact Info

Product

Resources

About