arXiv subjects

Kai Song

Publications and source records attributed to Kai Song.

34 records · Page 2Linked to original sources

arXiv2024

Semformer: Transformer Language Models with Semantic Planning

Next-token prediction serves as the dominant component in current neural language models. During the training phase, the model employs teacher forcing, which predicts tokens based on all preceding ground truth tokens. However, this approach has been found to create shortcuts, utilizing the revealed prefix to spuriously fit future tokens, potentially compromising the accuracy of the next-token predictor. In this paper, we introduce Semformer, a novel method of training a Transformer language model that explicitly models the semantic planning of response. Specifically, we incorporate a sequence of planning tokens into the prefix, guiding the planning token representations to predict the latent semantic representations of the response, which are induced by an autoencoder. In a minimal planning task (i.e., graph path-finding), our model exhibits near-perfect performance and effectively mitigates shortcut learning, a feat that standard training methods and baseline models have been unable to accomplish. Furthermore, we pretrain Semformer from scratch with 125M parameters, demonstrating its efficacy through measures of perplexity, in-context learning, and fine-tuning on summarization tasks.

cs.CL↗

arXiv2024

Adaptive Hybrid Masking Strategy for Privacy-Preserving Face Recognition Against Model Inversion Attack

The utilization of personal sensitive data in training face recognition (FR) models poses significant privacy concerns, as adversaries can employ model inversion attacks (MIA) to infer the original training data. Existing defense methods, such as data augmentation and differential privacy, have been employed to mitigate this issue. However, these methods often fail to strike an optimal balance between privacy and accuracy. To address this limitation, this paper introduces an adaptive hybrid masking algorithm against MIA. Specifically, face images are masked in the frequency domain using an adaptive MixUp strategy. Unlike the traditional MixUp algorithm, which is predominantly used for data augmentation, our modified approach incorporates frequency domain mixing. Previous studies have shown that increasing the number of images mixed in MixUp can enhance privacy preservation but at the expense of reduced face recognition accuracy. To overcome this trade-off, we develop an enhanced adaptive MixUp strategy based on reinforcement learning, which enables us to mix a larger number of images while maintaining satisfactory recognition accuracy. To optimize privacy protection, we propose maximizing the reward function (i.e., the loss function of the FR system) during the training of the strategy network. While the loss function of the FR network is minimized in the phase of training the FR network. The strategy network and the face recognition network can be viewed as antagonistic entities in the training process, ultimately reaching a more balanced trade-off. Experimental results demonstrate that our proposed hybrid masking scheme outperforms existing defense algorithms in terms of privacy preservation and recognition accuracy against MIA.

cs.CV↗

arXiv2024

Power Transformer Fault Prediction Based on Knowledge Graphs

In this paper, we address the challenge of learning with limited fault data for power transformers. Traditional operation and maintenance tools lack effective predictive capabilities for potential faults. The scarcity of extensive fault data makes it difficult to apply machine learning techniques effectively. To solve this problem, we propose a novel approach that leverages the knowledge graph (KG) technology in combination with gradient boosting decision trees (GBDT). This method is designed to efficiently learn from a small set of high-dimensional data, integrating various factors influencing transformer faults and historical operational data. Our approach enables accurate safe state assessments and fault analyses of power transformers despite the limited fault characteristic data. Experimental results demonstrate that this method outperforms other learning approaches in prediction accuracy, such as artificial neural networks (ANN) and logistic regression (LR). Furthermore, it offers significant improvements in progressiveness, practicality, and potential for widespread application.

cs.LG↗

arXiv2023

Single-pixel imaging based on deep learning

Single-pixel imaging can collect images at the wavelengths outside the reach of conventional focal plane array detectors. However, the limited image quality and lengthy computational times for iterative reconstruction still impede the practical application of single-pixel imaging. Recently, deep learning has been introduced into single-pixel imaging, which has attracted a lot of attention due to its exceptional reconstruction quality, fast reconstruction speed, and the potential to complete advanced sensing tasks without reconstructing images. Here, this advance is discussed and some opinions are offered. Firstly, based on the fundamental principles of single-pixel imaging and deep learning, the principles and algorithms of single-pixel imaging based on deep learning are described and analyzed. Subsequently, the implementation technologies of single-pixel imaging based on deep learning are reviewed. They are divided into super-resolution single-pixel imaging, single-pixel imaging through scattering media, photon-level single-pixel imaging, optical encryption based on single-pixel imaging, color single-pixel imaging, and image-free sensing according to diverse application fields. Finally, major challenges and corresponding feasible approaches are discussed, as well as more possible applications in the future.

eess.IV↗

arXiv2023

A Spectral Approach for the Dynamic Bradley-Terry Model

The dynamic ranking, due to its increasing importance in many applications, is becoming crucial, especially with the collection of voluminous time-dependent data. One such application is sports statistics, where dynamic ranking aids in forecasting the performance of competitive teams, drawing on historical and current data. Despite its usefulness, predicting and inferring rankings pose challenges in environments necessitating time-dependent modeling. This paper introduces a spectral ranker called Kernel Rank Centrality, designed to rank items based on pairwise comparisons over time. The ranker operates via kernel smoothing in the Bradley-Terry model, utilizing a Markov chain model. Unlike the maximum likelihood approach, the spectral ranker is nonparametric, demands fewer model assumptions and computations, and allows for real-time ranking. We establish the asymptotic distribution of the ranker by applying an innovative group inverse technique, resulting in a uniform and precise entrywise expansion. This result allows us to devise a new inferential method for predictive inference, previously unavailable in existing approaches. Our numerical examples showcase the ranker's utility in predictive accuracy and constructing an uncertainty measure for prediction, leveraging data from the National Basketball Association (NBA). The results underscore our method's potential compared to the gold standard in sports, the Arpad Elo rating system.

stat.ME↗

arXiv2023

R-PMAC: A Robust Preamble Based MAC Mechanism Applied in Industrial Internet of Things

This paper proposes a novel media access control (MAC) mechanism, called the robust preamble-based MAC mechanism (R-PMAC), which can be applied to power line communication (PLC) networks in the context of the Industrial Internet of Things (IIoT). Compared with other MAC mechanisms such as P-MAC and the MAC layer of IEEE1901.1, R-PMAC has higher networking speed. Besides, it supports whitelist authentication and functions properly in the presence of data frame loss. Firstly, we outline three basic mechanisms of R-PMAC, containing precise time difference calculation, preambles generation and short ID allocation. Secondly, we elaborate its networking process of single layer and multiple layers. Thirdly, we illustrate its robust mechanisms, including collision handling and data retransmission. Moreover, a low-cost hardware platform is established to measure the time of connecting hundreds of PLC nodes for the R-PMAC, P-MAC, and IEEE1901.1 mechanisms in a real power line environment. The experiment results show that R-PMAC outperforms the other mechanisms by achieving a 50% reduction in networking time. These findings indicate that the R-PMAC mechanism holds great potential for quickly and effectively building a PLC network in actual industrial scenarios.

cs.IT↗

arXiv2023

Audience Expansion for Multi-show Release Based on an Edge-prompted Heterogeneous Graph Network

In the user targeting and expanding of new shows on a video platform, the key point is how their embeddings are generated. It's supposed to be personalized from the perspective of both users and shows. Furthermore, the pursue of both instant (click) and long-time (view time) rewards, and the cold-start problem for new shows bring additional challenges. Such a problem is suitable for processing by heterogeneous graph models, because of the natural graph structure of data. But real-world networks usually have billions of nodes and various types of edges. Few existing methods focus on handling large-scale data and exploiting different types of edges, especially the latter. In this paper, we propose a two-stage audience expansion scheme based on an edge-prompted heterogeneous graph network which can take different double-sided interactions and features into account. In the offline stage, to construct the graph, user IDs and specific side information combinations of the shows are chosen to be the nodes, and click/co-click relations and view time are used to build the edges. Embeddings and clustered user groups are then calculated. When new shows arrive, their embeddings and subsequent matching users can be produced within a consistent space. In the online stage, posterior data including click/view users are employed as seeds to look for similar users. The results on the public datasets and our billion-scale data demonstrate the accuracy and efficiency of our approach.

cs.SI↗

arXiv2022

Enhanced Preamble Based MAC Mechanism for IIoT-oriented PLC Network

In this paper, we propose an enhanced preamble based media access control mechanism (E-PMAC), which can be applied in power line communication (PLC) network for Industrial Internet of Things (IIoT). We introduce detailed technologies used in E-PMAC, including delay calibration mechanism, preamble design, and slot allocation algorithm. With these technologies, E-PMAC is more robust than existing preamble based MAC mechanism (P-MAC). Besides, we analyze the disadvantage of P-MAC in multi-layer networking and design the networking process of E-PMAC to accelerate networking process. We analyze the complexity of networking process in P-MAC and E-PMAC and prove that E-PMAC has lower complexity than P-MAC. Finally, we simulate the single-layer networking and multi-layer networking of E-PMAC, P-MAC, and existing PLC protocol, i.e. , IEEE1901.1. The simulation results indicate that E-PMAC spends much less time in networking than IEEE1901.1 and P-MAC. Finally, with our work, a PLC network based on E-PMAC mechanism can be realized.

eess.SP↗

arXiv2019

Arithmetic addition of two integers by deep image classification networks: experiments to quantify their autonomous reasoning ability

The unprecedented performance achieved by deep convolutional neural networks for image classification is linked primarily to their ability of capturing rich structural features at various layers within networks. Here we design a series of experiments, inspired by children's learning of the arithmetic addition of two integers, to showcase that such deep networks can go beyond the structural features to learn deeper knowledge. In our experiments, a set of images is constructed, each image containing an arithmetic addition $n+m$ in its central area, and several classification networks are then trained over a subset of images, using the sum as the label. Tests on the excluded images show that, as the image set gets larger, the networks have well learnt the law of arithmetic additions so as to build up their autonomous reasoning ability strongly. For instance, networks trained over a small percentage of images can classify a big majority of the remaining images correctly, and many arithmetic additions involving some integers that have never been seen during the training can also be solved correctly by the trained networks.

cs.CV↗

arXiv2019

Code-Switching for Enhancing NMT with Pre-Specified Translation

Leveraging user-provided translation to constrain NMT has practical significance. Existing methods can be classified into two main categories, namely the use of placeholder tags for lexicon words and the use of hard constraints during decoding. Both methods can hurt translation fidelity for various reasons. We investigate a data augmentation method, making code-switched training data by replacing source phrases with their target translations. Our method does not change the MNT model or decoding algorithm, allowing the model to learn lexicon translations by copying source-side target words. Extensive experiments show that our method achieves consistent improvements over existing approaches, improving translation of constrained words without hurting unconstrained words.

cs.CL↗

arXiv2018

Exact generator and its high order expansions in the time-convolutionless generalized master equation: Applications to the spin-boson model and exictation energy transfer

The time-convolutionless (TCL) quantum master equation provides a powerful tool to simulate reduced dynamics of a quantum system coupled to a bath. The key quantity in the TCL master equation is the so-called kernel or generator, which describes effects of the bath degrees of freedom. Since the exact TCL generators are usually hard to calculate analytically, most applications of the TCL generalized master equation have relied on approximate generators using second and fourth order perturbative expansions. By using the hierarchical equation of motion (HEOM) and extended HEOM methods, we present a new approach to calculate the exact TCL generator and its high order perturbative expansions. The new approach is applied to the spin-boson model with different sets of parameters, to investigate the convergence of the high order expansions of the TCL generator. We also discuss circumstances where the exact TCL generator becomes singular for the spin-boson model, and a model of excitation energy transfer in the Fenna-Matthews-Olson complex.

quant-ph↗

arXiv2018

Identifying viruses from metagenomic data by deep learning

The recent development of metagenomic sequencing makes it possible to sequence microbial genomes including viruses in an environmental sample. Identifying viral sequences from metagenomic data is critical for downstream virus analyses. The existing reference-based and gene homology-based methods are not efficient in identifying unknown viruses or short viral sequences. Here we have developed a reference-free and alignment-free machine learning method, DeepVirFinder, for predicting viral sequences in metagenomic data using deep learning techniques. DeepVirFinder was trained based on a large number of viral sequences discovered before May 2015. Evaluated on the sequences after that date, DeepVirFinder outperformed the state-of-the-art method VirFinder at all contig lengths. Enlarging the training data by adding millions of purified viral sequences from environmental metavirome samples significantly improves the accuracy for predicting under-represented viruses. Applying DeepVirFinder to real human gut metagenomic samples from patients with colorectal carcinoma (CRC) identified 51,138 viral sequences belonging to 175 bins. Ten bins were associated with the cancer status, indicating their potential use for non-invasive diagnosis of CRC. In summary, DeepVirFinder greatly improved the precision and recall rates of viral identification, and it will significantly accelerate the discovery rate of viruses.

q-bio.GN↗

arXiv2018

Improved English to Russian Translation by Neural Suffix Prediction

Neural machine translation (NMT) suffers a performance deficiency when a limited vocabulary fails to cover the source or target side adequately, which happens frequently when dealing with morphologically rich languages. To address this problem, previous work focused on adjusting translation granularity or expanding the vocabulary size. However, morphological information is relatively under-considered in NMT architectures, which may further improve translation quality. We propose a novel method, which can not only reduce data sparsity but also model morphology through a simple but effective mechanism. By predicting the stem and suffix separately during decoding, our system achieves an improvement of up to 1.98 BLEU compared with previous work on English to Russian translation. Our method is orthogonal to different NMT architectures and stably gains improvements on various domains.

cs.CL↗

arXiv2017

Transformable Biomimetic Liquid Metal Chameleon

Liquid metal (LM) is of current core interest for a wide variety of newly emerging areas. However, the functional materials thus made so far by LM only could display a single silver-white appearance. Here in this study, the new conceptual colorful LM marbles working like transformable biomimetic chameleons were proposed and fabricated from LM droplets through encasing them with fluorescent nano-particles. We demonstrated that this unique LM marble can be manipulated into various stable magnificent appearances as one desires. And it can also splitt and merge among different colors. Such multifunctional LM chameleon is capable of responding to the outside electric-stimulus and realizing shape transformation and discoloration behaviors as well. Further more, the electric-stimuli has been disclosed to be an easy going way to trigger the release of nano/micro-particles from the LM. The present fluorescent biomimetic liquid metal chameleon is expected to offer important opportunities for diverse unconventional applications, especially in a wide variety of functional smart material and color changeable soft robot areas.

cond-mat.mtrl-sci↗

arXiv2015

Inference of Markovian Properties of Molecular Sequences from NGS Data and Applications to Comparative Genomics

Next Generation Sequencing (NGS) technologies generate large amounts of short read data for many different organisms. The fact that NGS reads are generally short makes it challenging to assemble the reads and reconstruct the original genome sequence. For clustering genomes using such NGS data, word-count based alignment-free sequence comparison is a promising approach, but for this approach, the underlying expected word counts are essential. A plausible model for this underlying distribution of word counts is given through modelling the DNA sequence as a Markov chain (MC). For single long sequences, efficient statistics are available to estimate the order of MCs and the transition probability matrix for the sequences. As NGS data do not provide a single long sequence, inference methods on Markovian properties of sequences based on single long sequences cannot be directly used for NGS short read data. Here we derive a normal approximation for such word counts. We also show that the traditional Chi-square statistic has an approximate gamma distribution, using the Lander-Waterman model for physical mapping. We propose several methods to estimate the order of the MC based on NGS reads and evaluate them using simulations. We illustrate the applications of our results by clustering genomic sequences of several vertebrate and tree species based on NGS reads using alignment-free sequence dissimilarity measures. We find that the estimated order of the MC has a considerable effect on the clustering results, and that the clustering results that use a MC of the estimated order give a plausible clustering of the species.

q-bio.GN↗

arXiv2012

Nucleating Acicular Ferrite with Galaxite (Al2O3MnO) and Manganese Oxide (MnO2) in FeMnAlC steels

Effectiveness of galaxite (Al2O3MnO) and manganese oxide (MnO2) in nucleating acicular ferrite was demonstrated in a steel with composition of Fe-13.92Mn-4.53Al-1.28Si-0.11C. The hot rolled and quenched steel had a duplex microstructure of delta-ferrite and partially transformed austenite. Thermal treatments were used to manipulate the concentration and type of oxides and the ferrite plate density was found to correlate with inclusions of low misfit. Both bainitic and acicular ferrite were observed and the prior austenite had a grain diameter of 16.5 micrometers. Upon tensile testing the retained austenite transformed to alpha-martensite. Ultimate tensile strength and elongation were 970 MPa and 40 percent. The yield strength was found to be dependent on the ferrite plate size, which varied linearly with the strength of the austenite. Ductility decreased as the strength disparity between delta-ferrite and deformation-induced alpha-martensite increased.

cond-mat.mtrl-sci↗