Article
Physics, Multidisciplinary
Yidong Liao, Daniel Ebler, Feiyang Liu, Oscar Dahlsten
Summary: The importance of neural network performance for tasks lies in the initial calibration and training of parameters, existing training methods have issues, proposing the use of quantum superposition of weight configurations can lead to high probability convergence towards the globally optimal solution.
NEW JOURNAL OF PHYSICS
(2021)
Article
Automation & Control Systems
Cong Liu, Qingtian Zeng, Long Cheng, Hua Duan, Jiujun Cheng
Summary: This article investigates the role of data in business process similarity measures, proposing a data-aware workflow net and three different perspectives of similarity measures. A methodology is introduced to help analysts apply these measures, and the effectiveness of these methods is evaluated through comparative experiments.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2022)
Article
Mathematics, Applied
Meng-Ze Lyu, Jin-Min Wang, Jian-Bing Chen
Summary: This paper investigates the closed-form solutions for the probability distribution of time-variant maximal value process (MVP) for certain classes of Markov processes, and establishes a unified Volterra integral equation for the evolution of cumulative distribution functions (CDFs) of time-variant MVP for general continuous Markov processes. Analytical or numerical solutions for time-variant MVP of special continuous Markov processes and compound Poisson processes are derived, and several examples are provided to demonstrate the effectiveness of the theoretical results.
COMMUNICATIONS IN NONLINEAR SCIENCE AND NUMERICAL SIMULATION
(2021)
Article
Engineering, Environmental
Jyoti Rani, Tapas Tripura, Hariprasad Kodamana, Souvik Chakraborty, Prakash Kumar Tamboli
Summary: This paper presents a probabilistic wavelet neural operator auto-encoder (PWNOAE) that aims to learn the distribution of multivariate process data and apply them for fault detection and isolation. The PWNOAE combines the integral kernel with wavelet transformation in a probabilistic fashion to learn the distribution of multivariate time series, and exploits wavelets to learn complex time-frequency characteristics underlying the datasets for fault detection and isolation.
PROCESS SAFETY AND ENVIRONMENTAL PROTECTION
(2023)
Article
Chemistry, Multidisciplinary
Martin Magdin, Juraj Benc, Stefan Koprda, Zoltan Balogh, Daniel Tucek
Summary: This paper compares the success rates of three different models of multilayer neural networks in the classification phase, using the EmguCV, ML.NET, and Tensorflow.Net libraries. The paper emphasizes the importance of choosing the right model that achieves the required accuracy with minimum training time and introduces an application that allows customization of parameters and integration into other applications.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Yitao Ren, Peiyang Jin, Yiyang Li, Keming Mao
Summary: This paper proposes an effective ghost module based spectral network for hyperspectral image classification. It adopts Ghost3D module to reduce model parameter size by generating redundant feature maps with linear transformation. Ghost2D module with channel-wise attention is used to explore informative spectral feature representation. Compared with existing methods, the proposed approach achieves superior performance on three hyperspectral image datasets with fewer sample labelling and less resource consumption.
IET IMAGE PROCESSING
(2023)
Article
Automation & Control Systems
Xiangxiang Huang, Xianping Guo, Xin Wen
Summary: This study discusses a two-person zero-sum game for finite-horizon semi-Markov processes, focusing on the probability that the total payoff exceeds a prescribed goal within a finite horizon. The study establishes the Shapley equation and proves the existence of a saddle point under certain conditions. Additionally, a value iterative algorithm is developed to compute an e-saddle point and approximate the game value through solving a series of matrix games. The construction of the e-saddle point and the convergence of the algorithm are also demonstrated. Furthermore, the application of the main results is illustrated using an example on an inventory system.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2023)
Article
Computer Science, Artificial Intelligence
Linquan Xu, Yuwen Chen, Songmei Lu, Kunhua Zhong, Yujie Li, Bin Yi
Summary: Anemia is significantly correlated with many diseases, but current testing methods are time-consuming, tedious, or prone to errors. This study proposes a self-supervised causal features method using actor-critical reinforcement learning to predict hemoglobin concentration, achieving high prediction performance with a short inference time, suitable for mobile deployment and health self-screening.
IET IMAGE PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
Nafiseh Zarei, Payman Moallem, Mohammadreza Shams
Summary: The position of vehicles is determined using an algorithm that includes two stages of detection and prediction. The algorithm is flexible to achieve the required accuracy and speed, and is robust against vehicle scale changes.
IET COMPUTER VISION
(2023)
Article
Mathematics
Peter Glynn, Yanlin Qu
Summary: This paper demonstrates that Harris recurrent Markov chains and processes can be defined as a class of Markov chains and processes in which there exists a random time T, at which the distribution of the chain or process does not depend on its initial condition. Notably, independence assumptions regarding the post-T process or T do not play a role in this characterization. Since Harris chains and processes are known to include infinite sequences of regeneration times with various independence properties, the existence of this single T implies the existence of infinitely many regeneration occurrence times.
Article
Neurosciences
Takuya Koumura, Hiroki Terashima, Shigeto Furukawa
Summary: By using a computational model, researchers found that human sensitivity to amplitude modulation in natural sounds may have emerged as a result of optimization for natural sound recognition and it is associated with neurophysiological similarity in the auditory brain regions.
JOURNAL OF NEUROSCIENCE
(2023)
Article
Biochemistry & Molecular Biology
Roberta Bardini, Alfredo Benso, Gianfranco Politano, Stefano Di Carlo
Summary: Ontogenesis is the process of organism development from early stages to maturity, involving the maintenance of homeostasis despite environmental changes. It relies on various levels of genetic and epigenetic regulation to create phenotypic diversity and complex cellular structures. The Nets-Within-Nets formalism can simulate these regulatory mechanisms and has been applied to model critical processes in ontogenesis, such as the specification of Vulval Precursor Cells in C.Elegans.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2021)
Article
Neurosciences
Adam Takacs, Annet Bluschke, Maximilian Kleimaker, Alexander Munchau, Christian Beste
Summary: This study examined the neurophysiological mechanisms underlying binding processes in actions using EEG recordings and various analyses. The findings suggest that the binding of action features modulates pre-motor processes preceding motor execution, but not motor execution processes themselves.
HUMAN BRAIN MAPPING
(2021)
Article
Computer Science, Artificial Intelligence
Zhaoqilin Yang, Gaoyun An, Ruichen Zhang, Zhenxing Zheng, Qiuqi Ruan
Summary: This paper proposes a novel two-stream inflated 3D ConvNet based on sparse regularization (SRI3D) for action recognition. The l(1) norm is embedded in the loss function to allow the network to learn the sparsity of the output. Experimental results show that SRI3D has a competitive advantage on Kinetics-400, Something-Something V2, UCF-101, and HMDB-51 compared to other state-of-the-art models.
IET IMAGE PROCESSING
(2023)
Article
Statistics & Probability
Vincent Liang, Konstantin Borovkov
Summary: We propose a discretization method using Markov chains and a Brownian bridge correction to estimate the curvilinear boundary crossing probabilities for diffusion processes. The method converges to the exact probabilities as the time grid gets finer. Numerical results demonstrate the convergence rate of the proposed method with the Brownian bridge correction is O(n(-2)) for C-2 boundaries and a uniform time grid with n steps.
JOURNAL OF APPLIED PROBABILITY
(2023)