(µ + λ) Evolution Strategy with Socio-cognitive Mutation

2024-07-24T01:26:37Z

题名: (µ + λ) Evolution Strategy with Socio-cognitive Mutation 作者: Aleksandra,Urbańczyk; Krzysztof,Kucaba; Mateusz,Wojtulewicz; Marek,Kisiel-Dorohinicki; Leszek,Rutkowski; Piotr,Duda; Janusz,Kacprzyk; Xin,Yao; Siang,Yew Chong; Aleksander,Byrski 摘要:

Socio‐cognitive computing is a paradigm developed for the last several years in our research group. It consists of introducing mechanisms inspired by inter‐individual learning and cognition into metaheuristics. Different ver‐ sions of the paradigm have been successfully applied in hybridizing Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), Genetic Algorithms, Differ‐ ential Evolution, and Evolutionary Multi‐agent System (EMAS) metaheuristics. In this paper, we have followed our previous experiences in order to propose a novel mutation based on socio‐cognitive mechanism and test it based on Evolution Strategy (ES). The newly constructed versions were applied to popular benchmarks and compared with their reference versions.

ADVANCING GROUNDWATER MODELLING USING DEEP LEARNING METHODS

2024-07-23T02:11:58Z

题名: ADVANCING GROUNDWATER MODELLING USING DEEP LEARNING METHODS 作者: 蔡和江摘要:

Advances in groundwater level modelling and time series deep learning (DL) techniques have progressed separately with limited integrations. Against the backdrop of the successful advancements in Artificial Intelligence (AI) over the past few decades, we are currently witnessing the accelerated adoption of cutting-edge intelligent technologies in the field of hydrology. However, the application of deep learning in groundwater research continues to face numerous challenges. For instance, despite its powerful nonlinear fitting capabilities, deep learning models are often criticized and questioned due to their black-box nature, which limits their ability to advance the development of science. This is especially apparent in groundwater level studies, where the intricate dynamics and unique environmental variables create a need for more tailored deep learning applications in groundwater level simulation, which prompts researchers to hold new expectations for the development of deep learning in groundwater level simulation. Within these expectations, there are two crucial issues that need to be addressed: 1) How to integrate deep learning algorithms with the specific physical processes involved in groundwater modelling? 2) How to illuminate the black box of deep learning models and enhance human understanding of groundwater dynamics? This thesis contributes to addressing these challenges by undertaking three research works that explore the application of two novel intelligent technologies in catchment-scale groundwater level simulation. These research works provide new insights and avenues for resolving the aforementioned challenges. Specifically, this thesis consists of three main topics: (1) Examining the impacts of region-averaged hydrometeorological and hydrogeological characteristics on improving the accuracy of groundwater level prediction using machine learning. (2) Embedding groundwater-related water balance mechanisms into recurrent deep learning methods for groundwater level simulation. (3) Introducing a new perspective from the decision-making procedure of deep learning models by state-of-the-art interpretable techniques to explain and understand extreme groundwater dynamics. The first topic of this thesis introduces a well-designed deep learning model for groundwater level simulation, and explore the statistical relationship between the model's performance, catchment characteristics, and groundwater dynamics, supported by an ample amount of data. This research summarizes the common characteristics of basins suitable for simulating groundwater level dynamics using deep learning, thereby deepening the understanding of the features associated with using deep learning to simulate groundwater levels. The focus of the second topic is to explore deep learning models constrained by physics laws for simulating groundwater dynamics. Formulas related to groundwater-related water balances are incorporated as additional algorithmic bases and constraints within deep learning models for groundwater level simulation. In this hybrid model, the combination of physical constraints and deep learning techniques enhances the model’s ability to comprehend the hydrogeological and hydrometeorological properties of the catchments, thereby improving the accuracy and generalization capability for predicting groundwater level. The focus of the third topic is to investigate hydrological insights related to groundwater from the perspective of deep learning models. Two state-of-the-art interpretability techniques for deep learning models are employed to analyse the underlying causes of groundwater drought events at different scales and seasons. This study integrated cutting-edge explainable DL methods into groundwater drought studies, thereby providing a new perspective for analysing the cause of drought events. It underscores the ability of explainable DL to deepen the understanding of hydrological phenomena, highlighting the imperative of synthesizing knowledge from various disciplines. While I carefully acknowledge the existing limitations of current algorithms, this study also reveals prospects for their future development. Overall, this thesis demonstrates the tremendous potential of utilizing deep learning techniques based on artificial neural networks to drive advancements in groundwater simulation. With thoughtful and innovative utilization of more intelligent technologies, it can be anticipated that significant strides in addressing the urgent groundwater challenges we are currently facing. By applying deep learning techniques, This thesis offers fresh insights and practical solutions for better understanding and managing groundwater resources, contributing to incremental advancements in the field.

阅读设备与产品种类对虚假评论感知可信度的影响

2024-07-22T01:35:27Z

题名: 阅读设备与产品种类对虚假评论感知可信度的影响作者: 原嘉宁摘要:

研究背景：当下电商平台虚假评论的泛滥对消费者，商家和平台都造成了损害。本研究从虚假评论的感知可信度出发，站在消费者的角度上研究阅读设备和产品种类作为语境因素的影响，进一步探索消费者与线上评论的互动机制。
研究内容：我们对语境因素进行进一步的定义和延伸，并通过解释水平理论研究了其中阅读设备和产品种类对虚假评论感知可信度的影响。我们认为阅读设备和产品种类作为语境因素影响了消费者阅读线上评论时的解释水平，而由于解释水平影响了消费者对于虚假评论的认知和判断，最终导致感知可信度发生变化。在本研究中，我们构建了两项实验来分别验证我们的假设。
研究结论：本研究通过实验发现，阅读设备和产品种类对虚假评论的感知可信度有着显著的影响，并且，当消费者使用智能手机作为阅读设备或购买搜寻品时，虚假评论的感知可信度相较于使用个人电脑作为阅读设备或购买经验品时要更高。但是，对于解释水平的中介效应，我们并没有发现显著的结果。
研究启示：本研究的结果让我们能够进一步了解消费者和线上评论的互动机制，基于在不同的语境因素下感知可信度的差异，平台可以采取针对性的政策来降低虚假评论的损害。

Dual hydration of oceanic lithospher (vol 10, nwad251, 2023)

2024-07-19T09:32:19Z

题名: Dual hydration of oceanic lithospher (vol 10, nwad251, 2023) 作者: Zhang, Fan; Lin, Jian; Zhu, Rixiang; Zhang, Xubo; Zhang, Jiangyang; Zhou, Zhiyuan

Influence of the Yucatan earthquake event (vol 11, 1201576, 2023)

2024-07-22T20:33:23Z

题名: Influence of the Yucatan earthquake event (vol 11, 1201576, 2023) 作者: Li, Changcheng

Identifying the wetlands of international importance in Beibu Gulf along the East Asian - Australasian Flyway, based on multiple citizen science datasets (vol 10, 1222806, 2023)

2024-07-22T20:33:23Z

题名: Identifying the wetlands of international importance in Beibu Gulf along the East Asian - Australasian Flyway, based on multiple citizen science datasets (vol 10, 1222806, 2023) 作者: Tang, Ningxin; Ma, Yanju; Li, Sixin; Yan, Yizhu; Cheng, Cheng; Lu, Gang; Li, Fei; Lv, Liuxuan; Qin, Peilin; Nguyen, Hoai Bao; Nguyen, Quang Hao; Le, Trong Trai; Wee, Shelby Qi Wei; He, Tao; Yong, Ding Li; Choi, Chi-Yeung

Application of single-cell sequencing to the research of tumor microenvironment (vol 14, 1285540, 2023)

2024-07-22T20:33:23Z

题名: Application of single-cell sequencing to the research of tumor microenvironment (vol 14, 1285540, 2023) 作者: Chen, Sijie; Zhou, Zhiqing; Li, Yu; Du, Yuhui; Chen, Guoan

Ionic migration induced loss analysis of perovskite solar cells: a poling study (vol 24, pg 7805, 2022)

2024-07-22T20:33:23Z

题名: Ionic migration induced loss analysis of perovskite solar cells: a poling study (vol 24, pg 7805, 2022) 作者: Zheng, Xue; Ming, Wenjie; Liu, Pingping; Zhang, Jie; Zhou, Hongfei; Chen, Ming; Li, Weimin; Huang, Boyuan; Wang, Huan; Yang, Chunlei 摘要: Correction for 'Ionic migration induced loss analysis of perovskite solar cells: a poling study' by Xue Zheng et al., Phys. Chem. Chem. Phys., 2022, 24, 7805-7814, https://doi.org/10.1039/D1CP05450C.

Nuclear export of circular RNA (vol 627, pg 212, 2024)

2024-07-22T20:33:23Z

题名: Nuclear export of circular RNA (vol 627, pg 212, 2024) 作者: Ngo, Linh H.; Bert, Andrew G.; Dredge, B. Kate; Williams, Tobias; Murphy, Vincent; Li, Wanqiu; Hamilton, William B.; Carey, Kirstyn T.; Toubia, John; Pillman, Katherine A.; Liu, Dawei; Desogus, Jessica; Chao, Jeffrey A.; Deans, Andrew J.; Goodall, Gregory J.; Wickramasinghe, Vihandha O.

A chest CT-based nomogram for predicting survival in acute myeloid leukemia ( vol 24 , 458 , 2024)

2024-07-22T20:33:23Z

题名: A chest CT-based nomogram for predicting survival in acute myeloid leukemia ( vol 24 , 458 , 2024) 作者: Yi, Xiaoping; Zhan, Huien; Lyu, Jun; Du, Juan; Dai, Min; Zhao, Min; Zhang, Yu; Zhou, Cheng; Xu, Xin; Fan, Yi; Li, Lin; Dong, Baoxia; Jiang, Xinya; Xiao, Zeyu; Zhou, Jihao; Zhao, Minyi; Zhang, Jian; Fu, Yan; Chen, Tingting; Xu, Yang; Tian, Jie; Liu, Qifa; Zeng, Hui