Spatio-temporal estimation plays a vital role in numerous scientific and engineering tasks, particularly for novel or unobserved locations lacking historical references. Many areas remain unobserved by sensors due to their non-core location or pending development status. The states of these areas can only be estimated through similar nodes in the geospace, rather than through historical data with temporal trends. Estimating these unobserved node states is crucial for city-wide spatio-temporal sensing and urban development, extending beyond simple point or block data imputation. In this study, we introduce a diffusion point process in high-frequency space to develop a robust spatio-temporal diffusion transformer for urban estimation where partial historical reference data is lacking. Our approach decomposes spatio-temporal data into high and low-frequency components through wavelet transform, and trains a diffusion model of spatial temporal data with a transformer that operates on high frequency signals. We incorporate low-frequency signals as diffusion conditions in the transformer architecture to capture overall spatio-temporal profiles and gradual trends. To enhance the learning of each step, we design an diffusion model featuring a spatio-temporal attention module that adaptively captures interdependencies between time and space. Extensive experiments across diverse domains including traffic, economics, and environment demonstrate that our method significantly outperforms state-of-the-art baselines.

Address parsing is an essential task that transforms natural language descriptions into standardized addresses, crucial for numerous urban applications. Existing methods struggle with ambiguous expressions, and even Large Language Models face challenges adapting to specialized domains with limited data. In this study, we focus on developing a robust framework to map diverse address descriptions into a unified semantic space of standardized addresses. We propose the Adaptive Weighted Learning-based Address Parsing (AWLAP) framework, which enhances parsing effectiveness through two key components: a multi-level constrained classifier that mines correlations between geographic entities across hierarchies, and an integrated discriminator that adaptively guides optimization based on parsing complexity. We evaluate the AWLAP using real data from JD Logistics and Point-of-Interest addresses. Extensive experiments comparing against state-of-the-art methods demonstrate AWLAP's effectiveness and robustness in address parsing. The proposed AWLAP framework has been successfully deployed as an address parsing service in practical applications.

Large Language Models (LLMs) have demonstrated amazing capabilities in language generation, comprehension, and knowledge reasoning. However, relying on a single model can result in biased and unstable outcomes in many tasks. Multi-model collaboration has been introduced to enhance reasoning abilities on various tasks, but obtaining the correct answer from multiple candidate responses remains a challenge. To address this issue, we propose multi-model brainstorming based on prompt. It incorporates different models into a group for brainstorming, to reach a consensus answer after multiple rounds of reasoning elaboration and re-inference. Our experiments on diverse datasets demonstrate that the brainstorming can substantially improve the effectiveness in logical reasoning. Further, we observe that two small-parameter models can achieve accuracy comparable to a larger-parameter model through brainstorming, presenting a novel approach for the distributed deployment of LLMs.

Efficient sensing is critical for Internet of Things (IoT) applications, such as environmental monitoring and traffic management, where high quality sensing data is essential for decision-making. Traditional sensing methods, however, are often plagued by high deployment costs and incomplete data coverage, significantly limiting their practicality. Despite recent progress, these methods continue to face challenges in maintaining data accuracy, ultimately degrading the Quality of Service (QoS) for IoT applications. To address these limitations, we propose ASSTG, a novel framework that combines an Active Sampling strategy with Spatio-Temporal Graph learning to enable efficient and accurate IoT sensing. At its core, AS-STG is designed to minimize the sampling cost while ensuring the accuracy of the data. The framework begins by analyzing historical data to determine the minimum sampling requirements for accurate inference in subsequent time slots. It then constructs a spatio-temporal graph to model the complex relationships between sensing grids, capturing both spatial and temporal dynamics. To supplement the spatio-temporal information and further optimize representations, we introduce two contrastive learning tasks. Leveraging the refined representation, AS-STG strategically selects informationrich regions for sampling, ensuring that even a sparse subset of samples can provide comprehensive coverage of the entire sensing area. Finally, AS-STG employs matrix completion techniques to reconstruct the complete sensing data from these sparse samples. Extensive experiments on real-world datasets demonstrate that AS-STG significantly outperforms baselines in terms of inference accuracy, cost-efficiency, and scalability. By effectively reducing sampling costs without compromising QoS, AS-STG offers a robust and scalable solution for dynamic IoT sensing systems.

Semi-supervised learning holds great promise for many real-world applications, due to its ability to leverage both unlabeled and expensive labeled data. However, most semi-supervised learning algorithms still heavily rely on the limited labeled data to infer and utilize the hidden information from unlabeled data. We note that any semi-supervised learning task under the self-training paradigm also hides an auxiliary task of discriminating label observability. Jointly solving these two tasks allows full utilization of information from both labeled and unlabeled data, thus alleviating the problem of over-reliance on labeled data. This naturally leads to a new generic and efficient learning framework without the reliance on any domain-specific information, which we call FlexSSL. The key idea of FlexSSL is to construct a semi-cooperative “game”, which forges cooperation between a main self-interested semi-supervised learning task and a companion task that infers label observability to facilitate main task training. We show with theoretical derivation of its connection to loss re-weighting on noisy labels. Through evaluations on a diverse range of tasks, we demonstrate that FlexSSL can consistently enhance the performance of semi-supervised learning algorithms.

Detecting anomalies in large complex systems is a critical and challenging task. The difficulties arise from several aspects. First, collecting ground truth labels or prior knowledge for anomalies is hard in real-world systems, which often lead to limited or no anomaly labels in the dataset. Second, anomalies in large systems usually occur in a collective manner due to the underlying dependency structure among devices or sensors. Lastly, real-time anomaly detection for high-dimensional data requires efficient algorithms that are capable of handling different types of data (i.e. continuous and discrete). We propose a correlation structure-based collective anomaly detection (CSCAD) model for high-dimensional anomaly detection problem in large systems, which is also generalizable to semi-supervised or supervised settings. Our framework utilize graph convolutional network combining a variational autoencoder to jointly exploit the feature space correlation and reconstruction deficiency of samples to perform anomaly detection. We propose an extended mutual information (EMI) metric to mine the internal correlation structure among different data features, which enhances the data reconstruction capability of CSCAD. The reconstruction loss and latent standard deviation vector of a sample obtained from reconstruction network can be perceived as two natural anomalous degree measures. An anomaly discriminating network can then be trained using low anomalous degree samples as positive samples, and high anomalous degree samples as negative samples. Experimental results on five public datasets demonstrate that our approach consistently outperforms all the competing baselines.

We study the problem of offline Imitation Learning (IL) where an agent aims to learn an optimal expert behavior policy without additional online environment interactions. Instead, the agent is provided with a supplementary offline dataset from suboptimal behaviors. Prior works that address this problem either require that expert data occupies the majority proportion of the offline dataset, or need to learn a reward function and perform offline reinforcement learning (RL) afterwards. In this paper, we aim to address the problem without additional steps of reward learning and offline RL training for the case when demonstrations contain a large proportion of suboptimal data. Built upon behavioral cloning (BC), we introduce an additional discriminator to distinguish expert and non-expert data. We propose a cooperation framework to boost the learning of both tasks, Based on this framework, we design a new IL algorithm, where the outputs of the discriminator serve as the weights of the BC loss. Experimental results show that our proposed algorithm achieves higher returns and faster training speed compared to baseline algorithms.

Purchase prediction is an essential task in both online and offline retail industry, especially during major shopping festivals, when strong promotion boosts consumption dramatically. It is important for merchants to forecast such surge of sales and have better preparation. This is a challenging problem, as the purchase patterns during shopping festivals are significantly different from usual cases and also rare in historical data. Most existing methods fail at this problem due to the extremely scarce data samples as well as the inability to capture the complex macroscopic spatio-temporal dependencies in a city. To address this problem, we propose the Spatio-Temporal Meta-learning Prediction (STMP) model for purchase prediction during shopping festivals. STMP is a meta-learning based spatio-temporal multi-task deep generative model. It adopts a meta-learning framework with few-shot learning capability to capture both spatial and temporal data representations. A generative component then uses the extracted spatio-temporal representation and input data to infer the prediction results. Extensive experiments demonstrate the meta-learning generalization ability of STMP. STMP outperforms baselines in all cases, which shows the effectiveness of our model.

Accurate network-wide traffic state estimation is vital to many transportation operations and urban applications. However, existing methods often suffer from the scalability issue when performing real-time inference at the city-level, or not robust enough under limited data. Currently, GPS trajectory data from probe vehicles has become a popular data source for many transportation applications. GPS trajectory data has large coverage area, which is ideal for network-wide applications, but also has the disadvantage of being sparse and highly heterogeneous among different time and locations. In this study, we focus on developing a robust and interpretable network-wide traffic state imputation framework using partially observed traffic information. We introduce a new learning strategy, called self-interested coalitional learning (SCL), which forges cooperation between a main self-interested semi-supervised learning task and a discriminator as a critic to facilitate main task training while providing interpretability on the results. In our detailed model, we use a temporal graph convolutional variational autoencoder (TG-VAE) as the reconstructor, which models the complex spatio-temporal pattern in data and solves the main traffic state imputation task. A discriminator is introduced to output interpretable imputation confidence on the estimated results and also help to enhance the performance of the reconstructor. The framework is evaluated using a large GPS trajectory dataset from taxis in Jinan, China. Extensive experiments against the state-of-the-art baselines demonstrate the effectiveness and robustness of the proposed method for network-wide traffic state estimation.

In order to address the conflict between resource- hungry mobile applications and resource-constrained mobile devices (MDs), mobile-edge computing (MEC), which offers cloud computing capabilities at the edge of networks in close proximity to the MDs, is envisioned to be a promising approach. However, existing mobile-edge computation offloading studies only took the resource allocation between the MDs and MEC servers into consideration, and ignored the resource allocation between MEC and centralized cloud computing servers. Moreover, current MEC Hosted Networks mostly adopt the networking technology integrating cellular and core networks, which has the shortcomings of single networking mode, high congestion, high latency and energy consumption. Toward this end, we provide in this paper an architecture of centralized cloud and distributed MEC over hybrid fiber-wireless network, which has the features of supporting diverse network techniques, easy expansibility, high capacity and reliability, low latency and energy consumption. The problem of cloud-MEC collaborative computation offloading is studied and an approximation collaborative computation offloading scheme is proposed as our solution. Numerical results corroborate the energy efficiency of our proposed collaborative scheme.