arXiv subjects

Boyi Hou

Publications and source records attributed to Boyi Hou.

7 recordsLinked to original sources

arXiv2026

Semimartingale Optimal Transport with Jumps: A General Framework and Equivalent Formulations

We study a semimartingale optimal transport (SOT) problem where the cost depends on the full differential characteristics, and the minimisation is over all semimartingale laws with given marginals whose absolutely continuous characteristics lie in a prescribed closed convex set. Under only minimal assumptions of measurability, convexity, and lower semicontinuity on the cost function, we prove the existence of an optimal plan for SOT and establish a Kantorovich-type duality without time-regularity conditions. We further prove that SOT admits four equivalent formulations: (i) a Kantorovich duality formulation, (ii) a viscosity solution formulation of the Hamilton--Jacobi--Bellman equation, (iii) a martingale solution formulation via Markovian projection, (iv) a PDE formulation via weak solutions of a non-local Fokker--Planck--Kolmogorov equation. The framework simultaneously generalises classical optimal transport, martingale optimal transport, Schr\"odinger bridge problems, and barycentric weak optimal transport.

math.PR↗

arXiv2025

Huawei Cloud Model-as-a-Service on the CloudMatrix384 SuperPod

Scaled-out MoE LLMs and scaled-up SuperPods create new systems challenges for production Model-as-a-Service (MaaS), requiring disaggregation, low-latency communication, and decentralized serving. This report presents xDeepServe, the production serving system behind Huawei Cloud's MaaS offering on CloudMatrix384, a 48-server SuperPod with 384 Ascend 910C chips connected by a high-bandwidth UB fabric and global shared memory. It serves models including DeepSeek, Kimi, GLM, Qwen, and MiniMax, among others. xDeepServe is built around Transformerless, a disaggregated execution architecture that decomposes transformer inference into modular units -- attention, feedforward, and MoE -- and supports disaggregated Prefill-Decode and MoE-Attention deployments. To enable disaggregation, we develop XCCL, a memory-semantic communication layer providing microsecond-level point-to-point and scalable all-to-all primitives, and we extend FlowServe with decentralized DP groups and techniques to mitigate stragglers and synchronization variance. In a peak decoding configuration, xDeepServe reaches 2400 tokens/s per Ascend 910C chip at ~50ms time-per-output-token (TPOT).

cs.DC↗

arXiv2019

Towards Interpretable and Learnable Risk Analysis for Entity Resolution

Machine-learning-based entity resolution has been widely studied. However, some entity pairs may be mislabeled by machine learning models and existing studies do not study the risk analysis problem -- predicting and interpreting which entity pairs are mislabeled. In this paper, we propose an interpretable and learnable framework for risk analysis, which aims to rank the labeled pairs based on their risks of being mislabeled. We first describe how to automatically generate interpretable risk features, and then present a learnable risk model and its training technique. Finally, we empirically evaluate the performance of the proposed approach on real data. Our extensive experiments have shown that the learning risk model can identify the mislabeled pairs with considerably higher accuracy than the existing alternatives.

cs.DB↗

arXiv2019

Gradual Machine Learning for Aspect-level Sentiment Analysis

The state-of-the-art solutions for Aspect-Level Sentiment Analysis (ALSA) were built on a variety of deep neural networks (DNN), whose efficacy depends on large amounts of accurately labeled training data. Unfortunately, high-quality labeled training data usually require expensive manual work, and may thus not be readily available in real scenarios. In this paper, we propose a novel solution for ALSA based on the recently proposed paradigm of gradual machine learning, which can enable effective machine labeling without the requirement for manual labeling effort. It begins with some easy instances in an ALSA task, which can be automatically labeled by the machine with high accuracy, and then gradually labels the more challenging instances by iterative factor graph inference. In the process of gradual machine learning, the hard instances are gradually labeled in small stages based on the estimated evidential certainty provided by the labeled easier instances. Our extensive experiments on the benchmark datasets have shown that the performance of the proposed solution is considerably better than its unsupervised alternatives, and also highly competitive compared to the state-of-the-art supervised DNN techniques.

cs.LG↗

arXiv2019

Gradual Machine Learning for Entity Resolution

Usually considered as a classification problem, entity resolution (ER) can be very challenging on real data due to the prevalence of dirty values. The state-of-the-art solutions for ER were built on a variety of learning models (most notably deep neural networks), which require lots of accurately labeled training data. Unfortunately, high-quality labeled data usually require expensive manual work, and are therefore not readily available in many real scenarios. In this paper, we propose a novel learning paradigm for ER, called gradual machine learning, which aims to enable effective machine labeling without the requirement for manual labeling effort. It begins with some easy instances in a task, which can be automatically labeled by the machine with high accuracy, and then gradually labels more challenging instances by iterative factor graph inference. In gradual machine learning, the hard instances in a task are gradually labeled in small stages based on the estimated evidential certainty provided by the labeled easier instances. Our extensive experiments on real data have shown that the performance of the proposed approach is considerably better than its unsupervised alternatives, and highly competitive compared to the state-of-the-art supervised techniques. Using ER as a test case, we demonstrate that gradual machine learning is a promising paradigm potentially applicable to other challenging classification tasks requiring extensive labeling effort.

cs.DB↗

arXiv2018

r-HUMO: A Risk-Aware Human-Machine Cooperation Framework for Entity Resolution with Quality Guarantees

Even though many approaches have been proposed for entity resolution (ER), it remains very challenging to find one with quality guarantees. To this end, we proposea risk-aware HUman-Machine cOoperation framework for ER, denoted by r-HUMO. Built on the existing HUMO framework, r-HUMO similarly enforces both precision and recall levels by partitioning an ER workload between the human and the machine. However, r-HUMO is the first solution to optimize the process of human workload selection from a risk perspective. It iteratively selects human workload based on real-time risk analysis on human-labeled results as well as prespecified machine metrics. In this paper,we first introduce the r-HUMO framework and then present the risk analysis technique to prioritize the instances for manual labeling. Finally,we empirically evaluate r-HUMO's performance on real data. Our extensive experiments show that r-HUMO is effective in enforcing quality guarantees,and compared with the state-of-the-art alternatives, it can achieve better quality control with reduced human cost.

cs.HC↗

arXiv2018

Improving Machine-based Entity Resolution with Limited Human Effort: A Risk Perspective

Pure machine-based solutions usually struggle in the challenging classification tasks such as entity resolution (ER). To alleviate this problem, a recent trend is to involve the human in the resolution process, most notably the crowdsourcing approach. However, it remains very challenging to effectively improve machine-based entity resolution with limited human effort. In this paper, we investigate the problem of human and machine cooperation for ER from a risk perspective. We propose to select the machine-labeled instances at high risk of being mislabeled for manual verification. For this task, we present a risk model that takes into consideration the human-labeled instances as well as the output of machine resolution. Finally, we evaluate the performance of the proposed risk model on real data. Our experiments demonstrate that it can pick up the mislabeled instances with considerably higher accuracy than the existing alternatives. Provided with the same amount of human cost budget, it can also achieve better resolution quality than the state-of-the-art approach based on active learning.

cs.DB↗