Searcharxiv⌕ Search

arXiv subjects

Li Guo

Publications and source records attributed to Li Guo.

At least 181 records · Page 10Linked to original sources

Renormalisation via locality morphisms

This is a survey on renormalisation in the locality setup highlighting the role that locality morphisms can play for renormalisation purposes. Having set up a general framework to build regularisation maps, we illustrate renormalisation by locality algebra homomorphisms on three examples, the renormalisation at poles of conical zeta functions, branched zeta functions and iterated integrals stemming from Kreimer's toy model.

math-ph↗

Inter-Patient ECG Classification with Convolutional and Recurrent Neural Networks

The recent advances in ECG sensor devices provide opportunities for user self-managed auto-diagnosis and monitoring services over the internet. This imposes the requirements for generic ECG classification methods that are inter-patient and device independent. In this paper, we present our work on using the densely connected convolutional neural network (DenseNet) and gated recurrent unit network (GRU) for addressing the inter-patient ECG classification problem. A deep learning model architecture is proposed and is evaluated using the MIT-BIH Arrhythmia and Supraventricular Databases. The results obtained show that without applying any complicated data pre-processing or feature engineering methods, both of our models have considerably outperformed the state-of-the-art performance for supraventricular (SVEB) and ventricular (VEB) arrhythmia classifications on the unseen testing dataset (with the F1 score improved from 51.08 to 61.25 for SVEB detection and from 88.59 to 89.75 for VEB detection respectively). As no patient-specific or device-specific information is used at the training stage in this work, it can be considered as a more generic approach for dealing with scenarios in which varieties of ECG signals are collected from different patients using different types of sensor devices.

eess.SP↗

Improving Knowledge Graph Embedding Using Simple Constraints

Embedding knowledge graphs (KGs) into continuous vector spaces is a focus of current research. Early works performed this task via simple models developed over KG triples. Recent attempts focused on either designing more complicated triple scoring models, or incorporating extra information beyond triples. This paper, by contrast, investigates the potential of using very simple constraints to improve KG embedding. We examine non-negativity constraints on entity representations and approximate entailment constraints on relation representations. The former help to learn compact and interpretable representations for entities. The latter further encode regularities of logical entailment between relations into their distributed representations. These constraints impose prior beliefs upon the structure of the embedding space, without negative impacts on efficiency or scalability. Evaluation on WordNet, Freebase, and DBpedia shows that our approach is simple yet surprisingly effective, significantly and consistently outperforming competitive baselines. The constraints imposed indeed improve model interpretability, leading to a substantially increased structuring of the embedding space. Code and data are available at https://github.com/iieir-km/ComplEx-NNE_AER.

cs.AI↗

Deformations and their controlling cohomologies of $\mathcal{O}$-operators

$\mathcal{O}$-operators are important in broad areas in mathematics and physics, such as integrable systems, the classical Yang-Baxter equation, pre-Lie algebras and splitting of operads. In this paper, a deformation theory of $\mathcal{O}$-operators is established in consistence with the general principles of deformation theories. On the one hand, $\mathcal{O}$-operators are shown to be characterized as the Maurer-Cartan elements in a suitable graded Lie algebra. A given $\mathcal{O}$-operator gives rise to a differential graded Lie algebra whose Maurer-Cartan elements characterize deformations of the given $\mathcal{O}$-operator. On the other hand, a Lie algebra with a representation is identified from an $\mathcal{O}$-operator $T$ such that the corresponding Chevalley-Eilenberg cohomology controls deformations of $T$, thus can be regarded as an analogue of the André-Quillen cohomology for the $\mathcal{O}$-operator. Thereafter, infinitesimal and formal deformations of $\mathcal{O}$-operators are studied. In particular, the notion of Nijenhuis elements is introduced to characterize trivial infinitesimal deformations. Formal deformations and extendibility of order $n$ deformations of an $\mathcal{O}$-operator are also characterized in terms of the new cohomology theory. Applications are given to deformations of Rota-Baxter operators of weight 0 and skew-symmetric $r$-matrices for the classical Yang-Baxter equation. For skew-symmetric $r$-matrices, there is an independent Maurer-Cartan characterization of the deformations as well as an analogue of the André-Quillen cohomology controlling the deformations, which turn out to be equivalent to the ones obtained as $\mathcal{O}$-operators associated to the coadjoint representations. Finally, infinitesimal deformations of skew-symmetric $r$-matrices and their corresponding triangular Lie bialgebras are studied.

math.QA↗

Modeling Text with Graph Convolutional Network for Cross-Modal Information Retrieval

Cross-modal information retrieval aims to find heterogeneous data of various modalities from a given query of one modality. The main challenge is to map different modalities into a common semantic space, in which distance between concepts in different modalities can be well modeled. For cross-modal information retrieval between images and texts, existing work mostly uses off-the-shelf Convolutional Neural Network (CNN) for image feature extraction. For texts, word-level features such as bag-of-words or word2vec are employed to build deep learning models to represent texts. Besides word-level semantics, the semantic relations between words are also informative but less explored. In this paper, we model texts by graphs using similarity measure based on word2vec. A dual-path neural network model is proposed for couple feature learning in cross-modal information retrieval. One path utilizes Graph Convolutional Network (GCN) for text modeling based on graph representations. The other path uses a neural network with layers of nonlinearities for image modeling based on off-the-shelf features. The model is trained by a pairwise similarity loss function to maximize the similarity of relevant text-image pairs and minimize the similarity of irrelevant pairs. Experimental results show that the proposed model outperforms the state-of-the-art methods significantly, with 17% improvement on accuracy for the best case.

cs.IR↗

Revealing two radio active galactic nuclei extremely near PSR J0437$-$4715

Newton's gravitational constant $G$ may vary with time at an extremely low level. The time variability of $G$ will affect the orbital motion of a millisecond pulsar in a binary system and cause a tiny difference between the orbital period-dependent measurement of the kinematic distance and the direct measurement of the annual parallax distance. PSR J0437$-$4715 is the nearest millisecond pulsar and the brightest at radio. To explore the feasibility of achieving a parallax distance accuracy of one light-year, comparable to the recent timing result, with the technique of differential astrometry, we searched for compact radio sources quite close to PSR J0437$-$4715. Using existing data from the Very Large Array and the Australia Telescope Compact Array, we detected two sources with flat spectra, relatively stable flux densities of 0.9 and 1.0 mJy at 8.4 GHz and separations of 13 and 45 arcsec. With a network consisting of the Long Baseline Array and the Kunming 40-m radio telescope, we found that both sources have a point-like structure and a brightness temperature of $\geq$10$^7$ K. According to these radio inputs and the absence of counterparts in the other bands, we argue that they are most likely the compact radio cores of extragalactic active galactic nuclei rather than Galactic radio stars. The finding of these two radio active galactic nuclei will enable us to achieve a sub-pc distance accuracy with the in-beam phase-referencing very-long-baseline interferometric observations and provide one of the most stringent constraints on the time variability of $G$ in the near future.

astro-ph.HE↗

Commutative modified Rota-Baxter algebras, shuffle products and Hopf algebras

In this paper, we begin a systematic study of modified Rota-Baxter algebras, as an associative analogue of the modified classical Yang-Baxter equation. We construct free commutative modified Rota-Baxter algebras by a variation of the shuffle product and describe the structure both recursively and explicitly. We then provide these algebras with a Hopf algebra structure by applying a Hochschild cocycle.

math.RA↗

Knowledge Graph Embedding with Iterative Guidance from Soft Rules

Embedding knowledge graphs (KGs) into continuous vector spaces is a focus of current research. Combining such an embedding model with logic rules has recently attracted increasing attention. Most previous attempts made a one-time injection of logic rules, ignoring the interactive nature between embedding learning and logical inference. And they focused only on hard rules, which always hold with no exception and usually require extensive manual effort to create or validate. In this paper, we propose Rule-Guided Embedding (RUGE), a novel paradigm of KG embedding with iterative guidance from soft rules. RUGE enables an embedding model to learn simultaneously from 1) labeled triples that have been directly observed in a given KG, 2) unlabeled triples whose labels are going to be predicted iteratively, and 3) soft rules with various confidence levels extracted automatically from the KG. In the learning process, RUGE iteratively queries rules to obtain soft labels for unlabeled triples, and integrates such newly labeled triples to update the embedding model. Through this iterative procedure, knowledge embodied in logic rules may be better transferred into the learned embeddings. We evaluate RUGE in link prediction on Freebase and YAGO. Experimental results show that: 1) with rule knowledge injected iteratively, RUGE achieves significant and consistent improvements over state-of-the-art baselines; and 2) despite their uncertainties, automatically extracted soft rules are highly beneficial to KG embedding, even those with moderate confidence levels. The code and data used for this paper can be obtained from https://github.com/iieir-km/RUGE.

cs.AI↗

An algebraic formulation of the locality principle in renormalisation

We study the mathematical structure underlying the concept of locality which lies at the heart of classical and quantum field theory, and develop a machinery used to preserve locality during the renormalisation procedure. Viewing renormalisation in the framework of Connes and Kreimer as the algebraic Birkhoff factorisation of characters on a Hopf algebra with values in a Rota-Baxter algebra, we build locality variants of these algebraic structures, leading to a locality variant of the algebraic Birkhoff factorisation. This provides an algebraic formulation of the conservation of locality while renormalising. As an application in the context of the Euler-Maclaurin formula on cones, we renormalise the exponential generating function which sums over the lattice points in convex cones. For a suitable multivariate regularisation, renormalisation from the algebraic Birkhoff factorisation amounts to composition by a projection onto holomorphic multivariate functions.

math-ph↗

Left counital Hopf algebra structures on free commutative Nijenhuis algebras

Motivated by the Hopf algebra structures established on free commutative Rota-Baxter algebras, we explore Hopf algebra related structures on free commutative Nijenhuis algebras. Applying a cocycle condition, we first prove that a free commutative Nijenhuis algebra on a left counital bialgebra (in the sense that the right-sided counicity needs not hold) can be enriched to a left counital bialgebra. We then establish a general result that a connected graded left counital bialgebra is a left counital right antipode Hopf algebra in the sense that the antipode is also only right-sided. We finally apply this result to show that the left counital bialgebra on a free commutative Nijenhuis algebra on a connected left counital bialgebra is connected and graded, hence is a left counital right antipode Hopf algebra.

math.RA↗

Rota-Baxter paired modules and their constructions from Hopf algebras

In this paper, we introduce the concept of a Rota-Baxter paired module to study Rota-Baxter modules without necessarily a Rota-Baxter operator. We obtain two characterizations of Rota-Baxter paired modules, and give some basic properties of Rota-Baxter paired modules. Beginning with the connection between the notion of integrals in the representations of Hopf algebra and of the notion of an integral algebra as a motivation and special case of Rota-Baxter algebra of weight zero, we obtain a large number of Rota-Baxter paired modules from Hopf related algebras and modules, including semisimple Hopf algebras, weak Hopf algebras, Long bialgebras, quasitriangular Hopf algebras, weak Hopf modules, dimodules and Doi-Hopf modules. Some of them give new examples of Rota-Baxter algebras.

math.QA↗

Rota-Baxter modules toward derived functors

In this paper we study Rota-Baxter modules with emphasis on the role played by the Rota-Baxter operators and resulting difference between Rota-Baxter modules and the usual modules over an algebra. We introduce the concepts of free, projective, injective and flat Rota-Baxter modules. We give the construction of free modules and show that there are enough projective, injective and flat Rota-Baxter modules to provide the corresponding resolutions for derived functor.

math.RA↗

On Rings of Differential Rota-Baxter Operators

Using the language of operated algebras, we construct and investigate a class of operator rings and enriched modules induced by a derivation or Rota-Baxter operator. In applying the general framework to univariate polynomials, one is led to the integro-differential analogs of the classical Weyl algebra. These are analyzed in terms of skew polynomial rings and noncommutative Groebner bases.

math.RA↗

Drazin inverses of the sum of four matrices and its applications

Our aim is to establish relations between Drazin inverses of the pesudo-block matrix $(P,Q,R,S)$ and the block matrix composed of $P,R,S,Q$, where $R^2=S^2=0$. Based on the relations, we give representations for Drazin inverses of the sum $P+Q+R+S$ under weaker restrictions. As its applications, several expressions for Drazin inverses of a $2\times 2$ block matrix are presented under some assumptions. Our results generalize several results in the literature.

math.RA↗

3-Lie bialgebras and 3-Lie classical Yang-Baxter equations in low dimensions

In this paper, we give some low-dimensional examples of local cocycle 3-Lie bialgebras and double construction 3-Lie bialgebras which were introduced in the study of the classical Yang-Baxter equation and Manin triples for 3-Lie algebras. We give an explicit and practical formula to compute the skew-symmetric solutions of the 3-Lie classical Yang-Baxter equation (CYBE). As an illustration, we obtain all skew-symmetric solutions of the 3-Lie CYBE in complex 3-Lie algebras of dimension 3 and 4 and then the induced local cocycle 3-Lie bialgebras. On the other hand, we classify the double construction 3-Lie bialgebras for complex 3-Lie algebras in dimensions 3 and 4 and then give the corresponding 8-dimensional pseudo-metric 3-Lie algebras.

math.QA↗

Weak composition quasi-symmetric functions, Rota-Baxter algebras and Hopf algebras

We introduce the Hopf algebra of quasi-symmetric functions with semigroup exponents generalizing the Hopf algebra QSym of quasi-symmetric functions. As a special case we obtain the Hopf algebra WCQSym of weak composition quasi-symmetric functions, which provides a framework for the study of a question proposed by G.-C.~Rota relating symmetric type functions and Rota-Baxter algebras. We provide the transformation formulas between the weak composition monomial and fundamental quasi-symmetric functions, which extends the corresponding results for quasi-symmetric functions. Moreover, we show that QSym is a Hopf subalgebra and a Hopf quotient algebra of WCQSym. Rota's question is addressed by identifying WCQsym with the free commutative unitary Rota-Baxter algebra of weight 1 on one generator, which also allows us to equip this algebra with a Hopf algebra structure.

math.CO↗

A conical approach to Laurent expansions for multivariate meromorphic germs with linear poles

We use convex polyhedral cones to study a large class of multivariate meromorphic germs, namely those with linear poles, which naturally arise in various contexts in mathematics and physics. We express such a germ as a sum of a holomorphic germ and a linear combination of special non-holomorphic germs called polar germs. In analyzing the supporting cones -- cones that reflect the pole structure of the polar germs -- we obtain a geometric criterion for the non-holomorphicity of linear combinations of polar germs. This yields the uniqueness of the above sum when required to be supported on a suitable family of cones and assigns a Laurent expansion to the germ. Laurent expansions provide various decompositions of such germs and thereby a uniformized proof of known results on decompositions of rational fractions. These Laurent expansions also yield new concepts on the space of such germs, all of which are independent of the choice of the specific Laurent expansion. These include a generalization of Jeffrey-Kirwan's residue, a filtered residue and a coproduct in the space of such germs. When applied to exponential sums on rational convex polyhedral cones, the filtered residue yields back exponential integrals.

math.CV↗

Quantified-Self 2.0: Using Context-Aware Services for Promoting Gradual Behaviour Change

The recent development of smartphone and wearable sensor technologies enable general public to carry self-tracking tasks more easily. Much work has been devoted to life data collection and visualisation to help people with better self-understanding. We believe that although (awareness/knowledge discovery is an important aspect of personal informatics, knowledge maintenance is more, or at least equally, important. In this paper, we propose a proactive approach that uses the knowledge mined from people's activity data to nudge them towards a good lifestyle (better knowledge maintenance). For demonstration purpose, a trial study was designed and implemented for good sleep maintenance. In the study, we first use smartphones as activity trackers to collect various features in a non-intrusive manner. We then use those data to learn users' activity patterns, including daily step amount, app usages, bedding time, wake-up time and sleep duration. Subsequently, we analyse correlations that may have the positive or negative impact on users' sleep qualities and finally we designed and implemented three proactive services that are able to generate customised advice in the "right" context to nudge users towards a better lifestyle. The experiments results are positive showing that with the use of the proposed services 1. daily step amount have been increased by 3.03% on average in a 10 days study and 2. sleep durations are increased by 7% for two subjects.

cs.CY↗