Review
Fisheries
K. J. Lees, M. A. MacNeil, K. J. Hedges, N. E. Hussey
Summary: Traditional mark-recapture methods in fisheries may have challenges and biases, while acoustic telemetry mark-recapture methods can overcome these issues but are not commonly used in fisheries management. Studies have identified various types of acoustic telemetry mark-recapture designs, highlighting their benefits and providing planning considerations for wider application.
REVIEWS IN FISH BIOLOGY AND FISHERIES
(2021)
Article
Mathematical & Computational Biology
Siyun Liu, Tao Yu
Summary: In this article, a method for density estimation of data with a mixture structure is proposed, which nonparametrically estimates component density functions through weighted kernel density estimation. Extensive simulation studies and real data examples demonstrate the superiority of the proposed method over existing methods in most cases.
STATISTICS IN MEDICINE
(2021)
Article
Computer Science, Interdisciplinary Applications
Shonosuke Sugasawa, Genya Kobayashi
Summary: This study proposes a robust fitting method for mixture models based on weighted complete estimating equations (WCE). The WCE introduces weights to the complete estimating equations to downweight the outliers automatically. A novel expectation-estimating-equation (EEE) algorithm is also developed to solve the WCE.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2022)
Article
Biology
Samira Abuelgasim Mohamed, Abdelmutalab G. A. Azrag, Francis Obala, Shepard Ndlela
Summary: This study developed models to predict the intricate details of T. absoluta pest's development, survival, and reproduction, and found that temperatures between 20-25 degrees C are ideal for its life cycle. This is important for developing pest management strategies and addressing global climate change.
Article
Computer Science, Software Engineering
Victoria L. Cooper, James C. Bieron, Pieter Peers
Summary: This article demonstrates a robust method for estimating the model parameters of a fully-linear data-driven BRDF model from a reflectance map under known natural lighting. By leveraging reflectance similarities within material classes and using a Gaussian mixture model to approximate the space of homogeneous BRDFs, the method approximates a solution per material class and selects the best solution. The efficacy and robustness of the method are demonstrated using the MERL BRDF database under various natural lighting conditions, and a proof-of-concept real-world experiment is provided.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2022)
Article
Engineering, Chemical
Mokhtar Said, Abdullah M. Shaheen, Ahmed R. Ginidi, Ragab A. El-Sehiemy, Karar Mahmoud, Matti Lehtonen, Mohamed M. E. Darwish
Summary: The paper highlights the successful extraction of parameters in photovoltaic models using the TFWO algorithm, presenting its effectiveness compared to other recent optimization techniques. The study demonstrates the accuracy and close approximation of the TFWO algorithm to experimental data.
Article
Computer Science, Interdisciplinary Applications
Haixu Wang, Jiguo Cao
Summary: ODE models are widely used to characterize dynamical systems, but the parameters are often unknown. This paper introduces a new method called pCODE, which can estimate the ODE parameters without the need for providing derivatives and Hessian matrices. The pCODE package is easy to understand and apply, and it also includes an online application.
Article
Ecology
Gioele Passoni, Tim Coulson, Francesca Cagnacci, Peter Hudson, Daniel R. Stahler, Douglas W. Smith, Shelly Lachish
Summary: A central debate in ecology has been the long-running discussion on the role of apex predators in affecting the abundance and dynamics of their prey. This study presents a bioenergetic mechanistic model of a tritrophic system and investigates the impact of wolf reintroduction on the system. The model reveals the important role of wolves in shifting the elk population from being food-limited to being predator-limited and stabilizing elk numbers.
Article
Transportation Science & Technology
Qixiu Cheng, Zhiyuan Liu, Jifu Guo, Xin Wu, Ram Pendyala, Baloka Belezamo, Xuesong (Simon) Zhou
Summary: The fluid-based queueing model is important for traffic flow modeling and state estimation. This paper proposes a spatial queue model for oversaturated traffic systems and demonstrates its effectiveness through empirical data.
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES
(2022)
Article
Robotics
James Mount, Ming Xu, Les Dawes, Michael Milford
Summary: This letter presents an unsupervised system that predicts the performance of a VPR algorithm using a limited number of analogues training images. By selecting the optimal operating point, the deployment capability of autonomous systems in real-world scenarios can be enhanced.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Medicine, General & Internal
Mika Kivimaki, Gill Livingston, Archana Singh-Manoux, Nina Mars, Joni V. Lindbohm, Jaana Pentti, Solja T. Nyberg, Matti Pirinen, Emma L. Anderson, Aroon D. Hingorani, Pyry N. Sipila
Summary: This study evaluated the clinical value of four widely used dementia risk scores in estimating 10-year dementia risk and found that these risk scores had high error rates, limiting their usefulness in targeting people for dementia prevention.
Article
Biochemistry & Molecular Biology
Zhanpeng Wang, Jiaping Wang, Michael Kourakos, Nhung Hoang, Hyong Hark Lee, Iain Mathieson, Sara Mathieson
Summary: Population genetics heavily relies on simulated data for validation, inference, and intuition, but simulated data often fail to mirror real genetic data, limiting the scope of methods. A novel approach, pg-gan, has been developed to estimate parameters in population genetic models and can adapt to any population's data, accurately recovering input parameters and recapitulating real data features.
MOLECULAR ECOLOGY RESOURCES
(2021)
Article
Computer Science, Artificial Intelligence
Mingyang Zhao, Xiaohong Jia, Lubin Fan, Yuan Liang, Dong-Ming Yan
Summary: A novel hierarchical Gaussian mixture model method is proposed for fitting ellipses in noisy and outliers-contained data, which significantly improves fitting accuracy and robustness, and speeds up the fitting process. Extensive experiments show that the method has high robustness against outliers and noise, high fitting accuracy, and improved performance compared to state-of-the-art methods.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2021)
Article
Multidisciplinary Sciences
Andrea Costa, Sebastiano Salvidio, Johannes Penner, Marco Basile
Summary: The time-for-space substitution (TSS) in N-mixture models allows estimating population abundance and trend of a single population without spatial replication. Simulation-based evaluation shows that TSS estimates are generally in good agreement with real abundance, with trend and abundance estimation mainly affected by detection probability and population size.
SCIENTIFIC REPORTS
(2021)
Article
Mathematics, Applied
Paul Hagemann, Sebastian Neumayer
Summary: This paper analyzes the properties of invertible neural networks, with a focus on controlling the Lipschitz constants of the inverse networks. Changing the latent distribution to a Gaussian mixture model resolves the issue of exploding Lipschitz constants and significantly improves sampling quality in multimodal applications, as confirmed by numerical simulations.