Article
Engineering, Geological
Wan-Ying Chien, Yu-Chen Lu, C. Hsein Juang, Jia-Jyun Dong, Wen-Yi Hung
Summary: This article compares two random field approaches for evaluating the liquefaction potential at a selected site. The chosen random field approach has an effect on the generation of the stratigraphic model and the uncertainty of the liquefaction potential index. The results show that the MRF approach generates more continuous strata and the CRF approach has a more uniform strata uncertainty.
ENGINEERING GEOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Paul Riverain, Simon Fossier, Mohamed Nadif
Summary: This paper proposes a probabilistic model based on the Latent Block Model (LBM) for co-clustering and improves clustering performance by introducing pairwise semi-supervision. The paper presents a general probabilistic framework for incorporating must link and cannot link relationships and provides two inference algorithms for count data. Extensive experiments on simulated data and real-world attributed networks confirm the importance of the approach and demonstrate the effectiveness of the algorithms.
Article
Computer Science, Information Systems
Yigen He, Xuesen Shi, Yongqing Wang
Summary: This study proposes a modified character string-based LZW algorithm to reduce compression time in the compression process of telemetry data. By designing coding principles, dictionary update rules, and search strategies, this algorithm can effectively reduce the number of dictionary searches.
Article
Health Care Sciences & Services
Maria Franco-Villoria, Massimo Ventrucci, Havard Rue
Summary: Bayesian disease mapping is useful in describing risk variation over time and space, but it requires addressing the challenge of interpreting random effect precision parameters. We propose a reparametrized variance partitioning model that enhances interpretability by balancing the contribution of main and interaction effects using a mixing parameter.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2022)
Article
Computer Science, Artificial Intelligence
Linyan Gu, Lihua Yang, Feng Zhou
Summary: This article investigates the approximation ability and representational efficiency of conditional restricted Boltzmann machines (CRBMs). By providing specially designed analysis tools, the results on the representational power of restricted Boltzmann machines (RBMs) are improved, and the universal approximation and maximal approximation error results for CRBMs are obtained. Additionally, the study discovers deterministic conditional distributions that cannot be computed by CRBMs.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Muhammad Hameed Siddiqi
Summary: This study focuses on improving the emotional speech classifier by introducing a novel methodology to address the limitations of existing classifiers, achieving significant improvement in emotional recognition. The proposed method has been validated and evaluated on two datasets, showing significantly improved classification performance. In terms of computation, the technique is also more cost-effective compared to state of the art works.
EGYPTIAN INFORMATICS JOURNAL
(2021)
Article
Parasitology
Mohamed F. F. Sallam, Shelley Whitehead, Narayani Barve, Amely Bauer, Robert Guralnick, Julie Allen, Yasmin Tavares, Seth Gibson, Kenneth J. J. Linthicum, Bryan V. V. Giordano, Lindsay P. P. Campbell
Summary: Mosquito vectors of EEEV and WNV in the USA vary in their composition and abundance, which affects pathogen transmission risk and vector control management. This study used CRF to examine spatial co-occurrence patterns between mosquito vectors and found that landscape and climate variables did not substantially improve the prediction of vector species abundance. The majority of vector species were positively dependent on other mosquito species, indicating that they may be habitat generalists.
PARASITES & VECTORS
(2023)
Article
Computer Science, Information Systems
Mingyang Li, Lin Shi, Yawen Wang, Junjie Wang, Qing Wang, Jun Hu, Xinhua Peng, Weimin Liao, Guizhen Pi
Summary: The paper introduces an automated approach named DEX for extracting data functions from textual requirements in Function Point Analysis (FPA), achieving promising results in evaluation on a real industrial dataset and manual review by domain experts. DEX outperforms state-of-the-art baselines and helps engineers produce more accurate and complete DFs in the industrial environment.
INFORMATION AND SOFTWARE TECHNOLOGY
(2022)
Article
Engineering, Electrical & Electronic
Zhen Guo, Zengfu Wang, Hua Lan, Quan Pan, Kun Lu
Summary: The study considers the spatial variation and correlation of the ionosphere, proposing a novel joint optimization solution to improve the accuracy of target localization by utilizing measurements from different sources.
Article
Engineering, Electrical & Electronic
Wenhai Qi, Mingxuan Sha, Ju H. Park, Huaicheng Yan, Xiangpeng Xie
Summary: This article focuses on the asynchronous stabilization of discrete hidden semi-Markov jumping power models subject to cyber attacks. A hidden semi-Markov model strategy is proposed to describe the mode mismatch between the original system and control law during operation. A new controller is constructed based on emission probability information, which relies on the observed mode rather than the system mode. Techniques for s-error mean square stability are developed using stochastic Lyapunov function under random deception attacks framework. Furthermore, an observed-mode-dependent controller is presented based on standard matrix inequality. An example is provided to demonstrate the effectiveness of the control scheme.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS
(2023)
Article
Optics
Xuanmin Zhu, Yuanchun Deng, Dezheng Zhang, Runping Gao, Qun Wei, Zijiang Luo
Summary: In this letter, the authors study the quantum search via continuous-time quantum walk on truncated simplex lattices, which is an important example of the quantum search on the structured database. They derive the run time of the quantum search and the critical jumping rates when the search target is a set of marked vertices. The letter also discusses the quantum search with partial information of the location of the marked vertex and proposes an optimized two-stage quantum search algorithm on the second-order truncated simplex lattice.
LASER PHYSICS LETTERS
(2023)
Article
Computer Science, Artificial Intelligence
Houcemeddine Filali, Karim Kalti
Summary: This paper proposes a method to combine local and global search strategies to solve the optimization problem of gray-scale image segmentation, and validates its effectiveness through experimentation. The parameterization issue of the solution is addressed through an empirical study, showing the impact of each parameter on the behavior of the proposed solution.
Article
Mathematics
Anatoliy Swishchuk, Nikolaos Limnios
Summary: This paper introduces controlled discrete-time semi-Markov random evolutions and their applications in various systems, providing dynamic principles and dynamic programming equations for the limiting processes. The rates of convergence in the limit theorems are also presented.
Review
Computer Science, Information Systems
Kamalika Bhattacharjee, Sukanta Das
Summary: This paper aims to search for and rank the so-called good generators by conducting a survey and empirical testing of pseudo-random number generators. Different types of generators are classified, and selected widely used ones are tested and ranked based on the results.
COMPUTER SCIENCE REVIEW
(2022)
Article
Statistics & Probability
Zhiyan Shi, Cong Liu, Yan Fan, Dan Bao, Yang Chen
Summary: We study a strong limit theorem of delayed sums for Markov chains in bi-infinite environments by constructing a sequence of random variables with parameters. As a result, we obtain some important strong limit properties, including the generalized conditional relative entropy for Markov chains in bi-infinite random environment. The obtained results generalize the findings of Liu et al. (2015).
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2022)
Article
Computer Science, Artificial Intelligence
Jianjuan Liang, Cuong Tuan Nguyen, Bilan Zhu, Masaki Nakagawa
PATTERN ANALYSIS AND APPLICATIONS
(2019)
Article
Computer Science, Artificial Intelligence
Cuong Tuan Nguyen, Hung Tuan Nguyen, Kazuhiro Mita, Masaki Nakagawa
PATTERN RECOGNITION LETTERS
(2019)
Article
Computer Science, Artificial Intelligence
Cuong Tuan Nguyen, Bipin Indurkhya, Masaki Nakagawa
INTERNATIONAL JOURNAL ON DOCUMENT ANALYSIS AND RECOGNITION
(2020)
Article
Computer Science, Artificial Intelligence
Cuong Tuan Nguyen, Vu Tran Minh Khuong, Hung Tuan Nguyen, Masaki Nakagawa
PATTERN RECOGNITION LETTERS
(2020)
Article
Computer Science, Artificial Intelligence
Kha Cong Nguyen, Cuong Tuan Nguyen, Masaki Nakagawa
PATTERN RECOGNITION LETTERS
(2020)
Article
Computer Science, Artificial Intelligence
Nam Tuan Ly, Cuong Tuan Nguyen, Masaki Nakagawa
PATTERN RECOGNITION LETTERS
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Hung Tuan Nguyen, Cuong Tuan Nguyen, Haruki Oka, Tsunenori Ishioka, Masaki Nakagawa
Summary: This research paper presents an experiment on automatically scoring handwritten descriptive answers, achieving high accuracy using deep neural networks and a pre-trained automatic scoring system. The results demonstrate the potential for further research on end-to-end automatic scoring of descriptive answers.
FRONTIERS IN HANDWRITING RECOGNITION, ICFHR 2022
(2022)
Proceedings Paper
Computer Science, Information Systems
Huy Quang Ung, Cuong Tuan Nguyen, Hung Tuan Nguyen, Masaki Nakagawa
Summary: This study proposes a method for computer-assisted marking by clustering online handwritten mathematical expressions to improve the efficiency and reliability of marking. Experimental results show that the method performs well in terms of accuracy and marking cost.
DOCUMENT ANALYSIS AND RECOGNITION - ICDAR 2021, PT II
(2021)
Proceedings Paper
Computer Science, Information Systems
Nam Tuan Ly, Hung Tuan Nguyen, Masaki Nakagawa
Summary: The paper proposes a model called 2D-SACRN for offline handwritten text recognition, which utilizes self-attention mechanism and recurrent encoder to achieve the recognition process from feature sequence to label probability sequence to final label sequence, and the experimental results show similar or even better accuracy compared to state-of-the-art models on all datasets.
DOCUMENT ANALYSIS AND RECOGNITION - ICDAR 2021, PT I
(2021)
Proceedings Paper
Computer Science, Information Systems
Trung Tan Ngo, Hung Tuan Nguyen, Masaki Nakagawa
Summary: This paper introduces an end-to-end attention-based neural network for identifying writers in historical documents. The model outperforms state-of-the-art results, showing promise in handling various sizes of historical document fragments for writer identification and image retrieval.
DOCUMENT ANALYSIS AND RECOGNITION - ICDAR 2021, PT II
(2021)
Proceedings Paper
Computer Science, Information Systems
Cuong Tuan Nguyen, Thanh-Nghia Truong, Hung Tuan Nguyen, Masaki Nakagawa
Summary: This paper introduces a temporal classification method for online handwritten mathematical expressions, trained by multiple paths of symbol and spatial relations derived from Symbol Relation Tree, benefiting from a deep bidirectional LSTM network for learning temporal classification. The method constructs a symbol-level parse tree with Context-Free Grammar to recognize online HME effectively.
DOCUMENT ANALYSIS AND RECOGNITION - ICDAR 2021, PT II
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Huy Quang Ung, Cuong Tuan Nguyen, Hung Tuan Nguyen, Thanh-Nghia Truong, Masaki Nakagawa
Summary: The paper introduces a Transformer-based Math Language Model to address ambiguities in handwritten mathematical expressions recognition. By training the model and incorporating it into the recognition system, significant improvements in expression rates have been achieved.
DOCUMENT ANALYSIS AND RECOGNITION, ICDAR 2021, PT II
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Trung Tan Ngo, Hung Tuan Nguyen, Nam Tuan Ly, Masaki Nakagawa
Summary: This paper proposes an RNN-Transducer model for recognizing Japanese and Chinese offline handwritten text line images, which combines visual and linguistic features and achieves state-of-the-art performance on two datasets through experiments.
DOCUMENT ANALYSIS AND RECOGNITION, ICDAR 2021, PT II
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Cuong Tuan Nguyen, Hung Tuan Nguyen, Kei Morizumi, Masaki Nakagawa
Summary: A temporal classification constraint was presented as an auxiliary learning method to improve recognition of Handwritten Mathematical Expression (HME), utilizing connectionist temporal classification (CTC) to learn temporal alignment of input feature sequence and corresponding symbol label sequence. Training CTC alignment through a combination of CTC loss and encoder-decoder loss was shown to enhance feature learning in the encoder of the encoder-decoder model, demonstrating effectiveness in symbol classification and expression recognition on the CROHME datasets.
DOCUMENT ANALYSIS AND RECOGNITION, ICDAR 2021, PT II
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Thanh-Nghia Truong, Huy Quang Ung, Hung Tuan Nguyen, Cuong Tuan Nguyen, Masaki Nakagawa
Summary: This paper proposes a relation-based sequence representation for offline handwritten mathematical expressions (HMEs) recognition, which outperforms the traditional LaTeX-based representation system. Experimental results show that the HME recognition system using the proposed representation achieves significantly higher recognition rates on the CROHME dataset.
DOCUMENT ANALYSIS AND RECOGNITION, ICDAR 2021 WORKSHOPS, PT I
(2021)