TFGAN: Traffic forecasting using generative adversarial network with multi-graph convolutional network

Article Computer Science, Artificial Intelligence

GSTA: gated spatial-temporal attention approach for travel time prediction

Alkilane Khaled et al.

Summary: This paper proposes a deep learning-based model, GSTA, for optimizing travel time prediction between two locations. Through a feature selection module and integrating multi-head attention, the model's performance is improved, as demonstrated in experiments showing its superiority over other models.

NEURAL COMPUTING & APPLICATIONS (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Bi-CLKT: Bi-Graph Contrastive Learning based Knowledge Tracing

Xiangyu Song et al.

Summary: The goal of Knowledge Tracing is to estimate students' mastery of a concept based on their learning history. With the rise of deep learning, Deep Knowledge Tracing has shown success in organizing students' learning plans.

KNOWLEDGE-BASED SYSTEMS (2022)

Add to Collection

Article Computer Science, Information Systems

JKT: A joint graph convolutional network based Deep Knowledge Tracing

Xiangyu Song et al.

Summary: Knowledge Tracing (KT) aims to track students' mastery of knowledge based on their historical learning interactions. To address the shortcomings of existing methods, a deep Knowledge Tracing framework called JKT is proposed, which utilizes a Joint graph convolutional network to model multi-dimensional relationships and improve interpretability.

INFORMATION SCIENCES (2021)

Add to Collection

Article Computer Science, Theory & Methods

Deep spatial-temporal sequence modeling for multi-step passenger demand prediction

Lei Bai et al.

Summary: The supply-demand imbalance poses challenges to transportation systems like taxis and shared vehicles, and accurate prediction of passenger demands is crucial. This study proposes an end-to-end deep learning framework to solve the prediction problem, which outperforms state-of-the-art methods by capturing spatial-temporal relationships effectively.

FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE (2021)

Add to Collection

Article Physics, Multidisciplinary

Hybrid time series forecasting methods for travel time prediction

Faruk Serin et al.

Summary: Providing accurate travel time information is crucial in public transportation. A novel three-layer architecture method is proposed to predict bus travel time between two stops, outperforming traditional approaches with an approximate MAPE of 6 in experiments using Istanbul's public transportation data.

PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS (2021)

Add to Collection

Article Engineering, Aerospace

Tensor completion-based trajectory imputation approach in air traffic control

Yi Lin et al.

Summary: This paper proposes a tensor completion-based approach to recover missing positions in air traffic trajectory data, achieving high accuracy imputation task through capturing global information and utilizing block coordinate descent algorithm.

AEROSPACE SCIENCE AND TECHNOLOGY (2021)

Add to Collection

Article Computer Science, Information Systems

Multi-Attention Generative Adversarial Network for Multivariate Time Series Prediction

Xiang Yin et al.

Summary: This paper proposes a novel multivariate time series prediction method based on multi-attention generative adversarial network, which includes three phases to explore prediction. Experimental results show that multi-attention mechanism is effective and the discrimination network can improve the performance in multivariate time series prediction.

IEEE ACCESS (2021)

Add to Collection

Article Computer Science, Information Systems

STI-GAN: Multimodal Pedestrian Trajectory Prediction Using Spatiotemporal Interactions and a Generative Adversarial Network

Lei Huang et al.

Summary: This paper introduces a method based on a generative adversarial network and a graph attention network for predicting the trajectories of multiple pedestrians in specific scenes. The method utilizes spatiotemporal interaction information to enhance trajectory prediction performance, reducing the average displacement error and final displacement error compared to existing methods.

IEEE ACCESS (2021)

Add to Collection

Article Engineering, Civil

TrafficGAN: Network-Scale Deep Traffic Prediction With Generative Adversarial Nets

Yuxuan Zhang et al.

Summary: Research on traffic flow prediction has been increasing, with a new model called TrafficGAN proposed for network-scale deep traffic prediction using Generative Adversarial Networks. By embedding CNN and LSTM models in TrafficGAN, along with a deformable convolution kernel for better data handling, significant improvements in short-term traffic flow prediction for road networks were achieved.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2021)

Add to Collection

Article Computer Science, Information Systems