检索结果-内蒙古大学图书馆

List-decoding of AG codes without genus penalty

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Beelen, Peter Montanucci, Maria Department of Applied Mathematics and Computer Science Technical University of Denmark Kongens Lyngby2800 Denmark

In this paper we consider algebraic geometry (AG) codes: a class of codes constructed from algebraic codes (equivalently, using function fields) by Goppa. These codes can be list-decoded using the famous Guruswami-Sudan (GS) list-decoder, but the genus g of the used function field gives rise to negative term in the decoding radius, which we call the genus penalty. In this article, we present a GS-like list-decoding algorithm for arbitrary AG codes, which we call the inseparable GS list-decoder. Apart from the multiplicity parameter s and designed list size , common for the GS list-decoder, we introduce an inseparability exponent e. Choosing this exponent to be positive gives rise to a list-decoder for which the genus penalty is reduced with a factor 1/pe compared to the usual GS list-decoder. Here p is the characteristic. Our list-decoder can be executed in Õ(sωµω−1pe(n + g)) field operations, where n is the code *** Codes 94B35, 94B27 Copyright © 2023, The Authors. All rights reserved.

关键词： Decoding

Availability and Demand of Jobs from the Perception of Automated Recruitment in Oman 7

学校读者我要写书评

暂无评论

Availability and Demand of Jobs from the Perception of Autom...

7th International Symposium on Multidisciplinary Studies and Innovative Technologies, ISMSIT 2023

作者： Karim Sadiq, Mohamed Abdul Al-Fannah, Nasser Mohammed Shanmugam, Thirumurugan Kausar, Firdous Oman Ministry of Transport Communications and Information Technology Muscat Oman College of Computing and Information Science Utas Suhar Oman College of Computing and Information Science University of Technology and Applied Sciences Suhar Oman Fisk University Department of Mathematics and Computer Science NashvilleTN United States

ISBN: (纸本)9798350342154

Natural Language Processing has been applied for automation of the recruitment process globally with the increase in volumes of profiles and jobs. This paper discusses certain challenges in the use of cutting-edge technologies for documents in the Arabic language. We analyze the availability and demand of jobs in Oman with reference to that in UAE. As a result, hurdles can be minimized in the job-resume matching process between the employers and prospective candidates. © 2023 IEEE.

关键词： Deep learning

On a tropicalization of planar polynomial ODEs with a finite number of structurally stable phase portraits

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Kristiansen, K.U. Sarantaris, A.H. Department of Applied Mathematics and Computer Science Technical University of Denmark Kgs Lyngby2800 Denmark

Recently, concepts from the emerging field of tropical geometry have been used to identify different scaling regimes in chemical reaction networks where dimension reduction may take place. In this paper, we try to formalize these ideas further in the context of planar polynomial ODEs. In particular, we develop a theory of a tropical dynamical system, based upon a differential inclusion, that has a set of discontinuities on a subset of the associated tropical curve. We define a phase portrait, a notion of equivalence and characterize structural stability for a tropical dynamical system. Our main result is that there are finitely many equivalence classes of structurally stable phase portraits. We believe that the property of finitely many structurally stable phase portraits underlines the potential of the tropical approach, also in higher dimension, as a method to (systematically) obtain and identify skeleton models in chemical reaction networks in extreme parameter regimes. Copyright © 2023, The Authors. All rights reserved.

关键词： Equivalence classes

WEIERSTRASS SEMIGROUPS AND AUTOMORPHISM GROUP OF A MAXIMAL CURVE WITH THE THIRD LARGEST GENUS

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Beelen, Peter Montanucci, Maria Vicino, Lara Department of Applied Mathematics and Computer Science Technical University of Denmark Kongens Lyngby2800 Denmark

In this article we explicitly determine the Weierstrass semigroup at any point and the full automorphism group of a known Fq2-maximal curve χ3 having the third largest genus. This curve arises as a Galois subcover of the Hermitian curve, and its uniqueness (with respect to the value of its genus) is a well-known open problem. Knowing the Weierstrass semigroups may provide a key towards solving this problem. Surprisingly enough χ3 has many different types of Weierstrass semigroups and the set of its Weierstrass points is much richer than the set of Fq2-rational points, as instead happens for all the known maximal curves where the Weierstrass points are known. We show that a similar exceptional behaviour does not occur in terms of automorphisms, that is, Aut(χ3) is exactly the automorphism group inherited from the Hermitian curve, apart from small values of *** Codes Primary: 11G20. Secondary: 11R58, 14H05, 14H55 Copyright © 2023, The Authors. All rights reserved.

关键词： Algebra

Discretization of the Cir Process Under Volatility Uncertainty Androbust Monte Carlo Simulation: An Application to Pricingguaranteed Equity-Linked Life Insurance Contract

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Akhtari, Bahar Payandeh Najafabadi, Amir T. Department of Applied Mathematics and Computer Science Faculty of Mathematics and Statistics University of Isfahan Isfahan81746-73441 Iran Department of Actuarial Science Faculty of Mathematical Sciences Shahid Beheshti University Tehran Iran

The non-negativity property of the CIR process driven by a G-Brownian motion makes it avaluable tool for term structure modeling in the fields of finance and insurance *** paper focuses on two specific types of guaranteed equity-linked contracts that are tiedto a financial asset whose price evolution follows the CIR process driven by G-Brownianmotion. The primary objective of this study is to assess the pricing of these contractsusing a robust expectation framework for the associated payoffs. To achieve this objective,we begin by developing a computationally efficient method to approximate the underlyingprocess while retaining key qualitative aspects of the model. We then demonstrate thenumerical convergence of this method. Subsequently, we introduce a robust Monte Carlotechnique that relies on a nite set of independent sample paths of the guaranteed payoffsat maturity to attain the desired outcome. © 2024, The Authors. All rights reserved.

关键词： Insurance

BRAIN-TO-TEXT DECODING WITH CONTEXT-AWARE NEURAL REPRESENTATIONS AND LARGE LANGUAGE MODELS

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Jingyuan Le, Trung Fan, Chaofei Chen, Mingfei Shlizerman, Eli Department of Electrical and Computer Engineering University of Washington United States Department of Computer Science Stanford University United States Department of Applied Mathematics Department of Electrical and Computer Engineering University of Washington United States

Decoding attempted speech from neural activity offers a promising avenue for restoring communication abilities in individuals with speech impairments. Previous studies have focused on mapping neural activity to text using phonemes as the intermediate target. While successful, decoding neural activity directly to phonemes ignores the context dependent nature of the neural activity-to-phoneme mapping in the brain, leading to suboptimal decoding performance. In this work, we propose the use of diphone - an acoustic representation that captures the transitions between two phonemes - as the context-aware modeling target. We integrate diphones into existing phoneme decoding frameworks through a novel divide-and-conquer strategy in which we model the phoneme distribution by marginalizing over the diphone distribution. Our approach effectively leverages the enhanced context-aware representation of diphones while preserving the manageable class size of phonemes, a key factor in simplifying the subsequent phoneme-to-text conversion task. We demonstrate the effectiveness of our approach on the Brain-to-Text 2024 benchmark, where it achieves state-of-the-art Phoneme Error Rate (PER) of 15.34% compared to 16.62% PER of monophone-based decoding. When coupled with finetuned Large Language Models (LLMs), our method yields a Word Error Rate (WER) of 5.77%, significantly outperforming the 8.93% WER of the leading method in the benchmark. Copyright © 2024, The Authors. All rights reserved.

关键词： Brain mapping

ProMIL: Probabilistic Multiple Instance Learning for Medical Imaging 26

学校读者我要写书评

暂无评论

ProMIL: Probabilistic Multiple Instance Learning for Medical...

26th European Conference on Artificial Intelligence, ECAI 2023

作者： Struski, Lukasz Rymarczyk, Dawid Lewicki, Arkadiusz Sabiniewicz, Robert Tabor, Jacek Zieliński, Bartosz Faculty of Mathematics and Computer Science Jagiellonian University Poland Faculty of Applied Computer Science University of Information Technology and Management Rzeszow Poland Department of Pediatric Cardiology and Congenital Heart Diseases Medical University of Gdansk Poland Ues Ltd. Doctoral School of Exact and Natural Sciences Jagiellonian University Poland

ISBN: (纸本)9781643684369

Multiple Instance Learning (MIL) is a weakly-supervised problem in which one label is assigned to the whole bag of instances. An important class of MIL models is instance-based, where we first classify instances and then aggregate those predictions to obtain a bag label. The most common MIL model is when we consider a bag as positive if at least one of its instances has a positive label. However, this reasoning does not hold in many real-life scenarios, where the positive bag label is often a consequence of a certain percentage of positive instances. To address this issue, we introduce a dedicated instance-based method called ProMIL, based on deep neural networks and Bernstein polynomial estimation. An important advantage of ProMIL is that it can automatically detect the optimal percentage level for decision-making. We show that ProMIL outperforms standard instance-based MIL in real-world medical applications. We make the code available. © 2023 The Authors.

关键词： Medical applications

Combination of Fuzzy Sets and Rough Sets for Machine Learning Purposes (Tutorial – Extended Abstract)

学校读者我要写书评

暂无评论

Combination of Fuzzy Sets and Rough Sets for Machine Learnin...

Federated Conference on computer science and Information Systems (FedCSIS)

作者： Chris Cornelis Henri Bollaert Department of Applied Mathematics Computational Web Intelligence Computer Science and Statistics Ghent University Belgium

Noting the highly complementary characteristics of fuzzy sets and rough sets, Dubois and Prade [7] proposed the first working definition of a fuzzy rough set, and thus paved the way for a flourishing hybrid theory wit...

关键词：

Extracting Key Information from Shopping Receipts by Using Bayesian Deep Learning via Multi-modal Features 3rd

学校读者我要写书评

暂无评论

Extracting Key Information from Shopping Receipts by Using...

3rd International Conference on Neural Computing for Advanced Applications, NCAA 2022

作者： Chen, Jiaqi Shao, Lujiao Zhou, Haibin Ma, Jianghong Meng, Weizhi Wang, Zenghui Zhang, Haijun Departmant of Computer Science Harbin Institudte of Technology Shenzhen China Department of Applied Mathematics and Computer Science Technical University of Denmark Kongens Lyngby Denmark Department of Electrical and Mining Engineering University of South Africa Florida1710 South Africa

ISBN: (纸本)9789811961410

This research presents a new key information extraction algorithm from shopping receipts. Specifically, we train semantic, visual and structural features through three deep learning methods, respectively, and formulate rule features according to the characteristics of shopping receipts. Then we propose a multi-class text classification algorithm based on multi-modal features using Bayesian deep learning. After post-processing the output of the classification algorithm, the key information we seek for can be obtained. Our algorithm was trained on a self-labeled Chinese shopping receipt dataset and compared with several baseline methods. Extensive experimental results demonstrate that the proposed method achieves optimal results on our Chinese receipt dataset. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Deep learning