Traffic prediction is essential for modern transportation systems, enhancing traffic management and urban planning. Accurate predictions of traffic flow and speed are crucial for understanding road usage, mitigating congestion, and providing real-time traffic monitoring and dynamic route guidance, thus improving road safety and infrastructure efficiency. Traditional research has often focused on predicting traffic flow or speed independently, leading to higher resource consumption due to the need for separate models. Few studies have explored the simultaneous prediction of both metrics, with recent attempts failing to account for spatial correlations, resulting in suboptimal performance. To address these challenges, we propose MTNet, a multi-task learning framework for joint traffic flow and speed prediction. MTNet employs a Transformer-like Encoder-Decoder architecture to process and enhance feature representations, capturing complex spatio-temporal correlations. Specifically, MTNet extracts intra-task dependencies using a cross-task interaction module and models task-specific spatiotemporal dependencies using spatial and temporal-aware modules with cascaded residual structures. Additionally, spatio-temporal positional encoding is integrated to increase awareness of long-term and long-distance dependencies. Extensive experiments on three diverse traffic datasets-Manchester, PeMSD4, and PeMSD8-demonstrate that MTNet significantly outperforms state-of-the-art methods in both traffic flow and speed prediction. MTNet achieves substantial improvements in prediction accuracy and efficiency, striking an optimal balance between performance and computational resource usage.

Characterizing post-translational modifications (PTMs) is crucial due to their vital role in regulating cellular activities. Current database search methods for peptide identification and PTM characterization face the challenge of exponentially increasing number of PTM combinations. Consequently, these methods resort to enumerating a small fraction of the PTM combinations. However, based on our analysis, more than 99% of the PTM combinations (up to three PTMs) are infeasible as their total mass does not match the mass shift between the precursor mass and theoretical mass. Enumerating those infeasible combinations is unnecessary and would lead to unsatisfactory performance. To address this issue, we propose a two-step method for characterizing multiple PTMs with a reduced search space of feasible PTM combinations only. The method first uses a mixed integer linear program (MILP) to collect feasible peptide sequences and PTM mass combinations. Then, these feasible peptides are exhaustively examined to find the optimal one. We conducted comprehensive experiments to evaluate the method’s performance. When applied to a soybean data set, our method has better performance than a representative tag-based search engine. The effectiveness of search space reduction and the sensitivity and precision of our method are demonstrated.

Traffic flow forecasting on a large-scale sensor network is of great practical significance for policy decision-making, urban management, and transport planning. Recently, several prediction methods based on graph convolution have been put forward. However, they are limited to small-scale analyses because of high computation complexity and fail to integrate spatial-temporal dependencies sufficiently between sensors with a large topological distance in multiple time steps. To address these issues, we propose a novel deep framework called PFNet to perform large-scale traffic forecasting. PFNet captures temporal correlations using deep multi-view sequence encoders (DMVSE) and spatial correlations using graph embedding technologies on transportation networks. Spatial-temporal dependencies are more comprehensively fused using cascaded progressive attention (CPA) modules due to the full use of latent temporal and spatial representation. Experiments are conducted on two large-scale traffic datasets (LondonHW and ManchesterHW) from England Highway. The results demonstrate that PFNet outperforms several existing state-of-the-art approaches.

Accurate spatio-temporal traffic forecasting serves as the basis of dynamic strategy and applications for intelligent transportation systems, which is of great practical significance for improving traffic safety and mitigating road congestion. Recently, deep learning methods such as convolutional neural networks (CNN) have been applied to traffic flow forecasting, which exhibits better performance than conventional methods. However, these CNN-based methods typically learn traffic as images to model spatial correlation, which is only applicable to Euclidean grid map data rather than non-Euclidean multi-sensor data. To address this problem, we propose a graph-based temporal attention framework GTA, which considers both spatial and temporal correlation, to forecast traffic flow based on data collected from multiple sensors. More specifically, GTA can better capture spatial dependencies leveraging graph embedding techniques on sensor networks because it preserves more details in the algorithms. We also introduce an attention mechanism to adaptively identify the relations among temporal submodules. Spatio-temporal dependencies are more effectively and comprehensively integrated due to the full use of the topological properties of transportation networks. We evaluate GTA with a large-scale traffic dataset from England and enhance it with topology information. The experimental results show that our approach outperforms several state-of-the-art baselines.

A novel framework for dynamic 3-D deployment of unmanned aerial vehicle (UAV) in the aerial base station system (ABSS) that based on the machine learning algorithms is proposed. In the framework, the UAV is deployed as an aerial base station to serve a group of ground users and is placed based on the prediction of the users' mobility. The joint problem of prediction of users' track and 3-D deployment of the UAV is formulated for maximizing the sum transmit rate. A two-step approach is proposed for predicting the movement of users and for determining the dynamic 3-D placement of the UAV. Firstly, an echo state network (ESN) based prediction algorithm is utilized for predicting the future positions of users based on the real-world datasets collected from Twitter. Secondly, an iterative K-Means based algorithm is proposed for obtaining the optimal placement of UAV at each time slot based on the output of ESN model. Numerical results are illustrated for showing the superiority of the proposed algorithm over the prevalent algorithm on prediction tasks. The accuracy and efficiency of the proposed framework are also investigated. Additionally, compared with static placement of the UAV, the advantage of dynamic 3-D deployment is demonstrated.

Recommender systems show increasingly importance with the development of E-commerce, news and multimedia applications. Traditional recommendation algorithms such as collaborative-filtering-based methods and graph-based methods mainly use items' original attributes and relationships between items and users, ignoring items' chronological order in browsing sessions. In recent years, RNN-based methods show their superiority when dealing with the sequential data, and some modified RNN models have been proposed. However, these RNN models only use the sequence order of items and neglect items' browsing time information. It is widely accepted that users tend to spend more time on their interested items, and these interested items are always closely related to users' current target. Based on the above view, items' browsing time is an important feature in recommendations. In this paper, we propose a modified RNN-based recommender system called TA4Rec, which can recommend the probable item that may be clicked in the next moment. Our main contribution is to introduce a method to calculate the time-attention factors from browsing items' duration time and add time-attention factors to the RNN-based model. We conduct experiments on RecSys Challenge 2015 dataset and the result shows that TA4REC model has gained obvious improvement on session-based recommendations than the classic session-based recommender method.