The Domain Generalizable Re-identification (DG ReID) task has attracted significant attention in recent years, as a challenging task but closely aligned with practical applications. Mixture-of-experts (MoE)-based methods have been studied for DG ReID to exploit the discrepancies and inherent correlations between diverse domains. However, most of DG ReID methods, especially MoE-based methods, have to fully fine-tune a large amount of parameters, which are not always practical in real-world scenarios. Considering this problem, we propose a novel MoE-based DG ReID method, named Mixture of Domain Adapters (MoDA), which utilizes many expert adapters and a global adapter to help MoE-based method scale to a much larger model but in a more parameter-efficient way. Furthermore, we conduct our approach with the large-scale vision-language pre-trained model CLIP, which exploits both visual and text encoders, to learn more robust representations based on multimodal information. Extensive experiments verify the effectiveness of our method and show that MoDA achieves competitiveness with state-of-the-art DG ReID methods with much fewer tunable parameters.

Multiaccess edge computing (MEC) has emerged as a promising solution that can enable low-end terminal devices to run large complex applications by offloading their tasks to edge servers. The task offloading strategy, determining how to offload tasks, remains the most critical issue of MEC. Traditional offloading approaches either suffer from high computational complexity or poor self-adjustability to dynamic changes in the edge environment. Deep reinforcement learning (DRL) provides an effective way to tackle these issues. However, most existing DRL-based methods solely consider either a continuous or a discrete action space, where the limited action space results in accuracy loss and restricts the optimality of offloading decisions. Nevertheless, the edge task offloading problem in practice often confronts both discrete and continuous actions. In this article, we propose a tailored proximal policy optimization (PPO)-based method, named Hybrid-PPO, enhanced by the parameterized discrete-continuous hybrid action space. Assisted with Hybrid-PPO, we further design a novel DRL-based multiserver multitask collaborative partial task offloading scheme adhering to a series of specifically built formal models. Experimental results prove that our approach achieves high offloading efficiency and outperforms the existing state-of-the-art offloading schemes in terms of convergence rate, energy cost, time cost, and generalizability under various network conditions.

Learning-to-rank is a core technique in the top-N recommendation task, where an ideal ranker would be a mapping from an item set to an arrangement (a.k.a. permutation). Most existing solutions fall in the paradigm of probabilistic ranking principle (PRP), i.e., first score each item in the candidate set and then perform a sort operation to generate the top ranking list. However, these approaches neglect the contextual dependence among candidate items during individual scoring, and the sort operation is non-differentiable. To bypass the above issues, we propose Set-To-Arrangement Ranking (STARank), a new framework directly generates the permutations of the candidate items without the need for individually scoring and sort operations; and is end-to-end differentiable. As a result, STARank can operate when only the ground-truth permutations are accessible without requiring access to the ground-truth relevance scores for items. For this purpose, STARank first reads the candidate items in the context of the user browsing history, whose representations are fed into a Plackett-Luce module to arrange the given items into a list. To effectively utilize the given ground-truth permutations for supervising STARank, we leverage the internal consistency property of Plackett-Luce models to derive a computationally efficient list-wise loss. Experimental comparisons against 9 the state-of-the-art methods on 2 learning-to-rank benchmark datasets and 3 top-N real-world recommendation datasets demonstrate the superiority of STARank in terms of conventional ranking metrics. Notice that these ranking metrics do not consider the effects of the contextual dependence among the items in the list, we design a new family of simulation-based ranking metrics, where existing metrics can be regarded as special cases. STARank can consistently achieve better performance in terms of PBM and UBM simulation-based metrics.

As an emerging field driven by industrial applications, integrated re-ranking combines lists from upstream sources into a single list, and presents it to the user. The quality of integrated re-ranking is especially sensitive to real-time user behaviors and preferences. However, existing methods are all built on the cloud-to-edge framework, where mixed lists are generated by the cloud model and then sent to the devices. Despite its effectiveness, such a framework fails to capture users' real-time preferences due to the network bandwidth and latency. Hence, we propose to place the integrated re-ranking model on devices, allowing for the full exploitation of real-time behaviors. To achieve this, we need to address two key issues: first, how to extract users' preferences for different sources from heterogeneous and imbalanced user behaviors; second, how to explore the correlation between the extracted personalized preferences and the candidate items. In this work, we present the first on-Device Integrated Re-ranking framework, DIR, to avoid delays in processing real-time user behaviors. DIR includes a multi-sequence behavior modeling module to extract the user's source-level preferences, and a preference-adaptive re-ranking module to incorporate personalized source-level preferences into the re-ranking of candidate items. Besides, we design exposure loss and utility loss to jointly optimize exposure fairness and overall utility. Extensive experiments on three datasets show that DIR significantly outperforms the state-of-the-art baselines in utility-based and fairness-based metrics.

User-tag profiling is an effective way of mining user attributes in modern recommender systems. However, prior researches fail to extract users' precise preferences for tags in the items due to their incomplete feature-input patterns. To convert user-item interactions to user-tag preferences, we propose a novel feature-based framework named Coalition Tag Multi-View Mapping (CTMVM), which identifies and investigates two special features, Coalition Feature and Privileged Feature. The former indicates decisive tags in each click where relationships between tags in one item are treated as a coalition game. The latter represents highly informative features that only occur during training. For the coalition feature, we adopt Shapley Value based Empowerment (SVE) to model the tags in items with a game-theoretic paradigm and charge the network to straight master user preferences for essential tags. For the privileged feature, we present Privileged Knowledge Mapping (PKM) to explicitly distill privileged feature knowledge for each tag into one single embedding, which assists the model in predicting user-tag preferences at a more fine-grained level. However, the barren capacity of single embeddings limits the diverse relations between each tag and different privileged features. Therefore, we further propose Adaptive Multi-View Mapping (AMVM) model to enhance effect by handling multiple mapping networks. Excellent offline experiment results on two public and one private datasets show the out-standing performance of CTMVM. After the deployment on Alibaba large-scale recommendation systems, CTMVM achieved improvement by 10.81% and 6.74% in terms of Theme-CTR and Item-CTR respectively, which validates the effectiveness of taking in the two particular features for training.

To ensure the delivery of high-performance and reliable services, data center networks (DCNs) are often over-provisioned for peak workload and traffic bursts. However, in real-world data centers, network traffic seldom reaches peak capacity of the network, resulting in significant energy waste. Traditional energy conservation approaches either suffer from high computational complexity and low solution quality, or their strategies cannot be dynamically adjusted to accommodate changes in data center network traffic. Deep reinforcement learning (DRL) provides an effective way to deal with these issues. However, most of the existing DRL-based schemes only consider either a continuous action space or a discrete action space, which greatly restricts the optimality of decisions. To solve these problems, this paper proposes a novel DRL-based DCN energy optimization framework, named SmartDCN. Specifically, SmartDCN consists of a traffic prediction module (TPM) and an energy optimization module (EOM). TPM incorporates an improved LSTM model JANET with an attention mechanism providing a high prediction accuracy, while EOM integrates our newly proposed parameterized DRL algorithm , named PAS-DQN, combining with the discrete-continuous hybrid action space. PAS-DQN implements a two-level control mechanism for the network, using TPM to predict future traffic in the data center as input. It is devoted to dynamically aggregating current traffic and makes tradeoffs between energy efficiency, performance, and robustness to optimize the network’s power consumption by dynamically calculating the minimum required network subset and turning off the non-involved network devices to achieve power savings. Experimental results show that SmartDCN significantly outperforms the existing state-of-the-art schemes in terms of energy savings under various network conditions.

In urban fields, mobile wireless sensor networks become ubiquitous. Accurate GPS positioning for sensors is a fundamental problem for MWSNs. This paper proposes a crowdsourcing aided positioning scheme to solve this problem. In crowd participant recruitment, two optimization objectives are addressed for the efficient aided positioning task. They are formulated as integer linear programming and their utility functions are proven to be submodular. Greedy algorithm is given for the two different optimizations. Simulation experiments are conducted to validate the algorithmic effectiveness at last.

With the popularity of cloud computing, many companies would outsource their social network data to a cloud service provider, where privacy leaks have become a more and more serious problem. However, most of the previous studies have ignored an important fact, i.e., in real social networks, users possess various attributes and have the flexibility to decide which attributes of their profiles are sensitive attributes by themselves. These sensitive attributes of the users should be protected from being revealed when outsourcing a social network to a cloud service provider. In this paper, we consider the problem of resisting privacy attacks with neighborhood information of both network structure and labels of one-hop neighbors as background knowledge. To tackle this problem, we propose a Global Similarity-based Group Anonymization (GSGA) method to generate a anonymized social network while maintaining as much utility as possible. We also extensively evaluate our approach on both real data set and synthetic data sets. Evaluation results show that the social network anonymized by our approach can still be used to answer aggregation queries with high accuracy.