Searcharxiv⌕ Search

arXiv subjects

Xiaomin Zhu

Publications and source records attributed to Xiaomin Zhu.

14 recordsLinked to original sources
arXiv2026

Observability from measurable sets for strongly coupled parabolic systems via single-component observation

We establish an observability inequality from space-time measurable sets for a class of strongly coupled parabolic systems consisting of two equations, where the observation acts on a single-component. The model is motivated by parabolic equations with complex coefficients and serves as a prototypical example of strongly coupled systems. The main difficulty lies in the fact that, unlike in the scalar and weakly coupled cases, pointwise-in-time interpolation observability estimates fail, as the observed component may exhibit high-frequency oscillatory cancellations induced by the coupling. To overcome this difficulty, we develop a new integral-type interpolation observability inequality based on a Remez-type inequality. With the aid of this integral-type interpolation observability inequality and the strategy developed in [Phung and Wang, JEMS, (2013), 681--703] and [Apraiz, Escauriaza,Wang and Zhang, JEMS, (2014), 2433--2475] for deriving observability from measurable sets, we obtain the desired observability inequality.

math.OC↗
arXiv2025

GENET: Unleashing the Power of Side Information for Recommendation via Hypergraph Pre-training

Recommendation with side information has drawn significant research interest due to its potential to mitigate user feedback sparsity. However, existing models struggle with generalization across diverse domains and types of side information. In particular, three challenges have not been addressed, and they are (1) the diverse formats of side information, including text sequences. (2) The diverse semantics of side information that describes items and users from multi-level in a context different from recommendation systems. (3) The diverse correlations in side information to measure similarity over multiple objects beyond pairwise relations. In this paper, we introduce GENET (Generalized hypErgraph pretraiNing on sidE informaTion), which pre-trains user and item representations on feedback-irrelevant side information and fine-tunes the representations on user feedback data. GENET leverages pre-training as a means to prevent side information from overshadowing critical ID features and feedback signals. It employs a hypergraph framework to accommodate various types of diverse side information. During pre-training, GENET integrates tasks for hyperlink prediction and self-supervised contrast to capture fine-grained semantics at both local and global levels. Additionally, it introduces a unique strategy to enhance pre-training robustness by perturbing positive samples while maintaining high-order relations. Extensive experiments demonstrate that GENET exhibits strong generalization capabilities, outperforming the SOTA method by up to 38% in TOP-N recommendation and Sequential recommendation tasks on various datasets with different side information.

cs.IR↗
arXiv2024

Recurrent Alignment with Hard Attention for Hierarchical Text Rating

While large language models (LLMs) excel at understanding and generating plain text, they are not tailored to handle hierarchical text structures or directly predict task-specific properties such as text rating. In fact, selectively and repeatedly grasping the hierarchical structure of large-scale text is pivotal for deciphering its essence. To this end, we propose a novel framework for hierarchical text rating utilizing LLMs, which incorporates Recurrent Alignment with Hard Attention (RAHA). Particularly, hard attention mechanism prompts a frozen LLM to selectively focus on pertinent leaf texts associated with the root text and generate symbolic representations of their relationships. Inspired by the gradual stabilization of the Markov Chain, recurrent alignment strategy involves feeding predicted ratings iteratively back into the prompts of another trainable LLM, aligning it to progressively approximate the desired target. Experimental results demonstrate that RAHA outperforms existing state-of-the-art methods on three hierarchical text rating datasets. Theoretical and empirical analysis confirms RAHA's ability to gradually converge towards the underlying target through multiple inferences. Additional experiments on plain text rating datasets verify the effectiveness of this Markov-like alignment. Our data and code can be available in https://github.com/ECNU-Text-Computing/Markov-LLM.

cs.CL↗
arXiv2024

A New Random Forest Ensemble of Intuitionistic Fuzzy Decision Trees

Classification is essential to the applications in the field of data mining, artificial intelligence, and fault detection. There exists a strong need in developing accurate, suitable, and efficient classification methods and algorithms with broad applicability. Random forest is a general algorithm that is often used for classification under complex conditions. Although it has been widely adopted, its combination with diverse fuzzy theory is still worth exploring. In this paper, we propose the intuitionistic fuzzy random forest (IFRF), a new random forest ensemble of intuitionistic fuzzy decision trees (IFDT). Such trees in forest use intuitionistic fuzzy information gain to select features and consider hesitation in information transmission. The proposed method enjoys the power of the randomness from bootstrapped sampling and feature selection, the flexibility of fuzzy logic and fuzzy sets, and the robustness of multiple classifier systems. Extensive experiments demonstrate that the IFRF has competitative and superior performance compared to other state-of-the-art fuzzy and ensemble algorithms. IFDT is more suitable for ensemble learning with outstanding classification accuracy. This study is the first to propose a random forest ensemble based on the intuitionistic fuzzy theory.

cs.AI↗
arXiv2019

An Automatic Design Framework of Swarm Pattern Formation based on Multi-objective Genetic Programming

Most existing swarm pattern formation methods depend on a predefined gene regulatory network (GRN) structure that requires designers' priori knowledge, which is difficult to adapt to complex and changeable environments. To dynamically adapt to the complex and changeable environments, we propose an automatic design framework of swarm pattern formation based on multi-objective genetic programming. The proposed framework does not need to define the structure of the GRN-based model in advance, and it applies some basic network motifs to automatically structure the GRN-based model. In addition, a multi-objective genetic programming (MOGP) combines with NSGA-II, namely MOGP-NSGA-II, to balance the complexity and accuracy of the GRN-based model. In evolutionary process, an MOGP-NSGA-II and differential evolution (DE) are applied to optimize the structures and parameters of the GRN-based model in parallel. Simulation results demonstrate that the proposed framework can effectively evolve some novel GRN-based models, and these GRN-based models not only have a simpler structure and a better performance, but also are robust to the complex and changeable environments.

cs.NE↗
arXiv2019

G-flocking: Flocking Model Optimization based on Genetic Framework

Flocking model has been widely used to control robotic swarm. However, with the increasing scalability, there exist complex conflicts for robotic swarm in autonomous navigation, brought by internal pattern maintenance, external environment changes, and target area orientation, which results in poor stability and adaptability. Hence, optimizing the flocking model for robotic swarm in autonomous navigation is an important and meaningful research domain.

cs.MA↗
arXiv2019

Not Just Privacy: Improving Performance of Private Deep Learning in Mobile Cloud

The increasing demand for on-device deep learning services calls for a highly efficient manner to deploy deep neural networks (DNNs) on mobile devices with limited capacity. The cloud-based solution is a promising approach to enabling deep learning applications on mobile devices where the large portions of a DNN are offloaded to the cloud. However, revealing data to the cloud leads to potential privacy risk. To benefit from the cloud data center without the privacy risk, we design, evaluate, and implement a cloud-based framework ARDEN which partitions the DNN across mobile devices and cloud data centers. A simple data transformation is performed on the mobile device, while the resource-hungry training and the complex inference rely on the cloud data center. To protect the sensitive information, a lightweight privacy-preserving mechanism consisting of arbitrary data nullification and random noise addition is introduced, which provides strong privacy guarantee. A rigorous privacy budget analysis is given. Nonetheless, the private perturbation to the original data inevitably has a negative impact on the performance of further inference on the cloud side. To mitigate this influence, we propose a noisy training method to enhance the cloud-side network robustness to perturbed data. Through the sophisticated design, ARDEN can not only preserve privacy but also improve the inference performance. To validate the proposed ARDEN, a series of experiments based on three image datasets and a real mobile application are conducted. The experimental results demonstrate the effectiveness of ARDEN. Finally, we implement ARDEN on a demo system to verify its practicality.

cs.LG↗
arXiv2018

Private Model Compression via Knowledge Distillation

The soaring demand for intelligent mobile applications calls for deploying powerful deep neural networks (DNNs) on mobile devices. However, the outstanding performance of DNNs notoriously relies on increasingly complex models, which in turn is associated with an increase in computational expense far surpassing mobile devices' capacity. What is worse, app service providers need to collect and utilize a large volume of users' data, which contain sensitive information, to build the sophisticated DNN models. Directly deploying these models on public mobile devices presents prohibitive privacy risk. To benefit from the on-device deep learning without the capacity and privacy concerns, we design a private model compression framework RONA. Following the knowledge distillation paradigm, we jointly use hint learning, distillation learning, and self learning to train a compact and fast neural network. The knowledge distilled from the cumbersome model is adaptively bounded and carefully perturbed to enforce differential privacy. We further propose an elegant query sample selection method to reduce the number of queries and control the privacy loss. A series of empirical evaluations as well as the implementation on an Android mobile device show that RONA can not only compress cumbersome models efficiently but also provide a strong privacy guarantee. For example, on SVHN, when a meaningful $(9.83,10^{-6})$-differential privacy is guaranteed, the compact model trained by RONA can obtain 20$\times$ compression ratio and 19$\times$ speed-up with merely 0.97% accuracy loss.

cs.LG↗
arXiv2018

Deep Learning Towards Mobile Applications

Recent years have witnessed an explosive growth of mobile devices. Mobile devices are permeating every aspect of our daily lives. With the increasing usage of mobile devices and intelligent applications, there is a soaring demand for mobile applications with machine learning services. Inspired by the tremendous success achieved by deep learning in many machine learning tasks, it becomes a natural trend to push deep learning towards mobile applications. However, there exist many challenges to realize deep learning in mobile applications, including the contradiction between the miniature nature of mobile devices and the resource requirement of deep neural networks, the privacy and security concerns about individuals' data, and so on. To resolve these challenges, during the past few years, great leaps have been made in this area. In this paper, we provide an overview of the current challenges and representative achievements about pushing deep learning on mobile devices from three aspects: training with mobile data, efficient inference on mobile devices, and applications of mobile deep learning. The former two aspects cover the primary tasks of deep learning. Then, we go through our two recent applications that apply the data collected by mobile devices to inferring mood disturbance and user identification. Finally, we conclude this paper with the discussion of the future of this area.

cs.LG↗
arXiv2015

Striking Dispersion of Recrystallized Poly(ethylene glycol)-Poly(lactic acid) Solvent-Casting Blend

The solvent casted Poly(ethylene glycol) (PEG) and Poly(lactic acid) (PLA) blend was recrystallized by the thermal treatment of a heating and cooling cycle. The high PEG content blend (30\%wt) showed a striking dispersion behavior, that the materials rapidly dispersed and dissolved in the aqueous environment in a few hours. This phenomenon has not been reported by others, and was not observed in low PEG content samples of 5\%, 10\%, and 20\%, or quenched(amorphous) samples. We hypothesized the mechanism that the chain rearrangement during the thermal treatment leads to the phase separation. And with the phase separation in the recrystallized samples, the PEG potions rapidly dissolved in the aqueous environment, left out the small PLA spherulites being separated and dispersed in the solution. The same underlying reason can also be inferred from the degradation behaviors of other samples. Characterizations of DSC, XRD, and SEM have been done to validate our hypothesis.

cond-mat.mtrl-sci↗
arXiv2014

The Crystallization Behavior of Porous PLA Prepared by Modified Solvent Casting/Particulate Leaching Technique for Potential Use of Tissue Engineering Scaffold

The porous PLA foams potential for tissue engineering usage are prepared by a modified solvent casting/particulate leaching method with different crystallinity. Since in typical method the porogens are solved in the solution and flow with the polymers during the casting and the crystallinity behavior of PLA chains in the limited space cannot be tracked, in this work the processing is modified by diffusing the PLA solution into a steady salt stack. With a thermal treatment before leaching while maintaining the stable structure of the porogens stack, the crystallinity of porous foams is made possible to control. The characterizations indicate the crystallization of porous foams is in a manner of lower crystallibility than the bulk materials. Pores and caves of around 250μm size are obtained in samples with different crystallinity. The macro-structures are not much impaired by the crystallization nevertheless the morphological effect of the heating process is still obvious.

cond-mat.mtrl-sci↗
arXiv2013

Loss Rate Based Fountain Codes for Data Transfer

Fountain codes are becoming increasingly important for data transferring over dedicated high-speed long-distance network. However, the encoding and decoding complexity of traditional fountain codes such as LT and Raptor codes are still high. In this paper, a new fountain codes named LRF (Loss Rate Based Fountain) codes for data transfer is proposed. In order to improve the performance of encoding and decoding efficiency and decrease the number of redundant encoding symbols, an innovative degree distribution instead of robust soliton degree distribution in LT (Luby Transfer) codes is proposed. In LRF codes, the degree of encoding symbol is decided by loss rate property, and the window size is extended dynamic. Simulations result using LRF codes show that the proposed method has better performance in term of encoding ratio, degree ratio, encoding and decoding efficiency with respect to LT and Raptor codes.

cs.NI↗