Article
Computer Science, Artificial Intelligence
Jincheng Li, Liangtu Song, Di Wu, Jiahao Shui, Tao Wang
Summary: Accurate financial time series forecasting is crucial in financial markets. However, a phenomenon called the lagging problem occurs in popular recurrent network models, where the predictive value lags behind the truth value for financial time series with low fluctuation. This study proposes new evaluation measures and analysis methods to explain the causes of the lagging problem, showing that popular recurrent network models suffer from it due to the failure of the nonlinear function and subsequent linear degeneration in the prediction model.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Chemistry, Analytical
Mingliang Yang, Kun Jiang, Junze Wen, Liang Peng, Yanding Yang, Hong Wang, Mengmeng Yang, Xinyu Jiao, Diange Yang
Summary: Deep neural network algorithms have achieved impressive performance in object detection. Real-time evaluation of perception uncertainty is crucial for safe driving in autonomous vehicles. This paper proposes a novel real-time evaluation method combining multi-source perception fusion and deep ensemble, and analyzes the spatial uncertainty of detected objects and its influencing factors.
Article
Thermodynamics
Subrata Bhattacharjee, Thomas Delzeit
Summary: In this study, closed-form expressions for the flame spread rate of opposed-flow flame spread over thermally thin and thick cylindrical fuels were developed using the scaling approach. A criterion for delineating thermally thin and thick cylinders was established based on the equality of the heated layer thickness and the radius. It was found that the spread rate of cylindrical fuels is always higher than that of flat fuels with a half-thickness equal to the radius of the cylindrical sample. For thermally thin cylinders, the predicted spread rate is at least two times faster than the corresponding flat fuel.
PROCEEDINGS OF THE COMBUSTION INSTITUTE
(2023)
Article
Computer Science, Interdisciplinary Applications
Kristian Fossum, Sergey Alyaev, Jan Tveranger, Ahmed H. Elsheikh
Summary: This paper proposes a method that uses GANs for parameterization and generation of geomodels, combined with EnRML for rapid updating of subsurface uncertainty. It illustrates the predictive ability of EnRML on assimilating well log data through several examples and verifies the results statistically using MCMC.
JOURNAL OF COMPUTATIONAL SCIENCE
(2022)
Article
Meteorology & Atmospheric Sciences
Anumeha Dube, S. Karunasagar, Raghavendra Ashrit, Ashis K. Mitra
Summary: This study uses an object-based spatial verification method to evaluate ensemble rainfall forecasts over three monsoon seasons in India and analyzes the contribution of displacement, volume, and pattern errors to the total error. The study finds that the uncertainties in the NEPS system are better represented in some regions and rainfall intensity is a challenging attribute to predict.
ATMOSPHERIC RESEARCH
(2022)
Article
Computer Science, Information Systems
Wu Chen, Yong Yu, Keke Gai, Jiamou Liu, Kim-Kwang Raymond Choo
Summary: The study introduced a decentralized framework called Multi-Agent Ensemble, leveraging edge computing to improve ensemble learning techniques by balancing access restrictions and accuracy enhancement. Experimental evaluations showed that Multi-Agent Ensemble outperforms other ensemble methods in terms of accuracy.
ACM TRANSACTIONS ON INTERNET TECHNOLOGY
(2021)
Article
Environmental Sciences
Dazhi Yang, Christian A. Gueymard
Summary: This paper aims to merge five gridded products of monthly aerosol optical depth using various regression techniques to generate probabilistic predictive distributions, improving uncertainty quantification compared to traditional techniques. The study compares the effectiveness of different merging techniques using distribution-oriented verification approach and finds that quantile regression forest is the best method for reducing root mean square error and bias in the final AOD product.
ATMOSPHERIC ENVIRONMENT
(2021)
Article
Physics, Multidisciplinary
L-L Yan, J-W Zhang, M-R Yun, J-C Li, G-Y Ding, J-F Wei, J-T Bu, B. Wang, L. Chen, S-L Su, F. Zhou, Y. Jia, E-J Liang, M. Feng
Summary: Dissipation is crucial in cyclic processes in realistic systems, and recent research on nonequilibrium processes in stochastic systems has revealed a dissipation-time uncertainty relation that restricts the evolution pace of physical processes. The researchers experimentally verified this relation and obtained the first experimental evidence confirming the thermodynamic restriction on quantum operations due to dissipation.
PHYSICAL REVIEW LETTERS
(2022)
Article
Geosciences, Multidisciplinary
Xu-Feng Yan, Hui Xu, Heng Lu, Jia-Wen Zhou, Xie-Kang Wang, Lu Wang
Summary: Global climate change and human activities are impacting the frequency and abruptness of water-related natural disasters in mountainous areas. A recent study of a debris flow event showed that accumulative wood and sediment deposition due to hillside vegetation formed a small-scale debris dam, leading to a subsequent debris flow.
FRONTIERS IN EARTH SCIENCE
(2021)
Article
Geosciences, Multidisciplinary
Sebastian Buschow, Petra Friederichs
Summary: Verification of high-resolution meteorological models requires highly resolved validation data and appropriate analysis tools. This study demonstrates how clear-sky radar echoes can provide insights into horizontal wind patterns, and how a two-dimensional wavelet transform can help quantify divergences and reflectivities in terms of their spatial scale, anisotropy, and direction. Despite some model shortcomings, the validation against radar data shows close agreement in spatial scales and wind direction.
GEOSCIENTIFIC MODEL DEVELOPMENT
(2021)
Article
Engineering, Civil
Velpuri Manikanta, Jew Das, K. Nikhil Teja, N. V. Umamahesh
Summary: This study examines the capability of ensemble precipitation forecasts obtained from two NWP models, and post-processes the raw ensemble members using two methods. The findings suggest that QRF post-processed forecasts are superior to other forecasts in terms of all verification measures at shorter lead times. However, the skill of both raw and post-processed forecasts declines at higher lead times.
JOURNAL OF HYDROLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Hadrien Bride, Cheng-Hao Cai, Jie Dong, Jin Song Dong, Zhe Hou, Seyedali Mirjalili, Jing Sun
Summary: This paper introduces a new high-performance machine learning tool named Silas, which aims to provide a more transparent, reliable, and efficient data analytics service. The advantages of Silas in predictive and computational performance are demonstrated, and customized algorithms in Silas outperform the state-of-the-art in terms of predictions in significantly shorter time. Furthermore, Silas focuses on providing a formal foundation of decision trees to support logical analysis and verification of learned prediction models.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Interdisciplinary Applications
Seungmin Yoo, Junho Song
Summary: This paper proposes a new method that combines computational wildfire simulations and ensemble Kalman filter for rapid prediction of wildfire spread. The method uses a two-dimensional polyline simplification algorithm to represent the wildfire perimeter and relates the prediction results with actual observation data. The proposed method is tested and demonstrated using an example wildfire spread scenario, showing that it can reduce computational time while maintaining prediction accuracy. It is expected to be a core algorithm for near-real-time prediction and data-driven updating of wildfire spread.
ENVIRONMENTAL MODELLING & SOFTWARE
(2023)
Article
Computer Science, Artificial Intelligence
Tuanfei Zhu, Cheng Luo, Zhihong Zhang, Jing Li, Siqi Ren, Yifu Zeng
Summary: This paper introduces a structure-preserving Oversampling method for high-dimensional imbalanced time series classification, OHIT, and integrates it into boosting framework to form a new ensemble algorithm OHITBoost. Extensive experiments on several publicly available time-series datasets demonstrate their effectiveness.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Environmental Sciences
Martina Idzanovic, Edel S. U. Rikardsen, Johannes Rohrs
Summary: The operational ocean Ensemble Prediction System for the coastal seas off Northern Norway is evaluated using high-frequency radar current speed estimates. The EPS consists of 24 members that are forced with an atmosphere ensemble instead of perturbed or constrained. The ensemble effectively predicts forecast uncertainty and demonstrates low but significant predictive skill in surface currents.
FRONTIERS IN MARINE SCIENCE
(2023)
Article
Meteorology & Atmospheric Sciences
Susanna Hagelin, Joohyung Son, Richard Swinbank, Anne McCabe, Nigel Roberts, Warren Tennant
QUARTERLY JOURNAL OF THE ROYAL METEOROLOGICAL SOCIETY
(2017)
Article
Astronomy & Astrophysics
E. Masciadri, J. Stoesz, S. Hagelin, F. Lascaux
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
(2010)
Article
Astronomy & Astrophysics
F. Lascaux, E. Masciadri, S. Hagelin
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
(2011)
Article
Astronomy & Astrophysics
F. Lascaux, E. Masciadri, S. Hagelin
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
(2010)
Article
Astronomy & Astrophysics
S. Hagelin, E. Masciadri, F. Lascaux
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
(2010)
Article
Astronomy & Astrophysics
S. Hagelin, E. Masciadri, F. Lascaux
MONTHLY NOTICES OF THE ROYAL ASTRONOMICAL SOCIETY
(2011)
Article
Meteorology & Atmospheric Sciences
Ludovic Auger, Olivier Dupont, Susanna Hagelin, Pierre Brousseau, Pascal Brovelli
QUARTERLY JOURNAL OF THE ROYAL METEOROLOGICAL SOCIETY
(2015)
Article
Meteorology & Atmospheric Sciences
Susanna Hagelin, Ludovic Auger, Pascal Brovelli, Olivier Dupont
WEATHER AND FORECASTING
(2014)
Article
Meteorology & Atmospheric Sciences
Aurore N. Porson, Susanna Hagelin, Douglas F. A. Boyd, Nigel M. Roberts, Rachel North, Stuart Webster, Jeff Chun-Fung Lo
QUARTERLY JOURNAL OF THE ROYAL METEOROLOGICAL SOCIETY
(2019)
Article
Meteorology & Atmospheric Sciences
Susanna Hagelin, Roohollah Azad, Magnus Lindskog, Harald Schyberg, Heiner Kornich
Summary: The impact of using wind observations from the Aeolus satellite in a limited-area NWP system over the Nordic region was investigated. The study found that Aeolus observations have a significant impact on the NWP model's analysis, but the quality of observations degraded over time. Adjusting the observation error for Aeolus data and utilizing 4D-Var data assimilation show potential improvements in the analysis.
ATMOSPHERIC MEASUREMENT TECHNIQUES
(2021)
Article
Meteorology & Atmospheric Sciences
Evgenia Belova, Sheila Kirkwood, Peter Voelger, Sourav Chatterjee, Karathazhiyath Satheesan, Susanna Hagelin, Magnus Lindskog, Heiner Kornich
Summary: This study compared winds measured by the Aeolus satellite with ground-based radars in Antarctica and Arctic Sweden. The results showed very good agreement between most data subsets, but identified potential biases in certain cases.
ATMOSPHERIC MEASUREMENT TECHNIQUES
(2021)
Article
Meteorology & Atmospheric Sciences
Evgenia Belova, Peter Voelger, Sheila Kirkwood, Susanna Hagelin, Magnus Lindskog, Heiner Kornich, Sourav Chatterjee, Karathazhiyath Satheesan
Summary: The ESRAD radar tends to underestimate zonal and meridional winds, likely due to receiver group arrangement and high non-white noise levels. In contrast, the MARA radar shows good consistency with radiosonde wind measurements, with minimal differences between the two datasets.
ATMOSPHERIC MEASUREMENT TECHNIQUES
(2021)