检索结果-内蒙古大学图书馆

Detection of outliers with respect to a MUSIC geotechnical database

Canadian Geotechnical Journal 2024年第7期61卷 1275-1293页

作者： Ching, Jianye Phoon, Kok-Kwang Huang, Pengsheng Department of Civil Engineering National Taiwan University Taipei Taiwan Information Systems Technology and Design/Architecture and Sustainable Design Singapore University of Technology and Design Singapore

This study proposes a novel method that addresses a nontraditional class of outlier detection problems. The purpose of most outlier detection methods in the literature is to detect outliers within a dataset. A record can be considered an outlier if it is distinct from the regular records in the dataset. However, the purpose of the novel outlier detection method proposed in this study is to detect outlier data groups (a data group may denote a site or a project) with respect to a soil/rock property "MUSIC" database. A data group is an outlier group if its characteristics (mean, variance, correlation, or higher order dependency) are distinct from the regular data groups in the database. This study frames the outlier detection problem into a formal hypothesis testing problems with the null hypothesis that "the target data group is identically distributed as the regular groups in the database." With the hierarchical Bayesian model previously developed by the first two authors, the p-value for this hypothesis testing problem can be estimated rigorously. Numerical and real examples show that the p-value can effectively detect outlier data groups as well as outlier records with respect to a database. © 2024 The Author(s).

关键词： Database systems

来源：评论

学校读者我要写书评

暂无评论

Editorial for Machine learning and AI for underground metaverse

引用

Underground Space 2024年第6期19卷 1-3页

作者： Kok-Kwang Phoon Qiujing Pan Chong Tang Information Systems Technology and Design/Architecture and Sustainable Design Singapore University of Technology and DesignSingapore 487372 School of Civil Engineering Central South UniversityChangsha 410075China State Key Laboratory of Coastal and Offshore Engineering Dalian University of TechnologyDalian 116024China

It is widely recognized that civil engineering practice,including underground construction,is lagging the fast pace digital transformation taking place in many *** key reason for this revolution is the emergence and convergence of digital technologies that include machine learning,artificial intelligence(AI),blockchain,cloud/edge computing,Internet of Things,5G communication,automation,digital twin,and *** parallel,Industry 4.0 is now complemented by Industry 5.0 that advocates a shift from economic value to societal value(sustainability,well-being,etc.).This means a shift from enterprises benefiting shareholders(profits)to stakeholders(people).

关键词： learning artificial underground

来源：评论

学校读者我要写书评

暂无评论

Transforming Large-Scale Participation Data Through Topic Modelling in Urban design Processes 20th

Transforming Large-Scale Participation Data Through Topic Mo...

引用

20th International Conference on Computer-Aided Architectural design Futures, CAAD Futures 2023

作者： Ataman, Cem Tunçer, Bige Perrault, Simon Architecture and Sustainable Design Singapore University of Technology and Design Singapore Singapore Information Systems Technology and Design Singapore University of Technology and Design Singapore Singapore

ISBN: (纸本)9783031371882

The advancements in digital tools and data collection methods ensure the continuing growth of textual data obtained through large-scale participation processes in urban contexts. In order to extract the thematic content of such underutilized textual datasets, topic modeling (TM) and content analysis have been deployed as promising AI-based Natural Language Processing (NLP) techniques. Yet, implementing such techniques has not been exploited in urban design domains due to the complexity of textual datasets and the lack of a systematic evaluation framework. In this paper, we addressed the challenges in the utilization of large textual data by using a real-world dataset collected via a digital participation platform in Madrid, Spain. Firstly, we identified prominent data structures and potential information embedded into the dataset by using a document-oriented NoSQL database. In this step, we systematically discussed data pre-processing steps to convert them into a series of structured data collections. Secondly, we evaluated three different TM algorithms, i.e. LDA, LSI, and HDP, according to a number of hyperparameters controlling the learning process. This step aimed to reveal the required number of topics to extract meaningful content through the algorithms. Lastly, we presented possible textual data visualization techniques to enable the use of textual information in digital participation processes. Consequently, this paper facilitates the use of large textual datasets by investigating data structures & processing, revealing the potentials of different TM algorithms, and eventually analyzing the results with the support of urban big data analytics and computational linguistic techniques for informed urban design processes. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Data Analytics

来源：评论

学校读者我要写书评

暂无评论

Bayesian Estimation for Subsurface Models using Spike-and-Slab Prior 8th

Bayesian Estimation for Subsurface Models using Spike-and-Sl...

引用

8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022

作者： Shuku, T. Phoon, K.K. Graduate School of Environment and Life Science Okayama University Japan Architecture and Sustainable Design/Information Systems Technology and Design Singapore University of Technology Singapore

ISBN: (纸本)9789811851841

A machine learning method called sparse modelling has received attention in many different fields. One of the method for achieving sparse modelling includes least absolute shrinkage and selection operator (lasso), which has been used for property estimation and layer boundary detection in subsurface modelling. The estimates by lasso, however, are point estimates and do not provide information on the "uncertainties." This study newly developed a Bayesian method for one-dimensional subsurface modelling using a spike-andslab prior. The proposed method was demonstrated through numerical examples on 1D subsurface modelling. The proposed method consistently solved two problems, property estimation and layer boundary detection, with high accuracy and can evaluate the posterior PDF of estimates based on Bayes’ rule. © 2022 ISRERM Organizers. Published by Research Publishing, Singapore.

关键词： Miocene

来源：评论

学校读者我要写书评

暂无评论

3D Data-driven Site Characterization using Geotechnical Lasso with Basis Functions 8th

3D Data-driven Site Characterization using Geotechnical Lass...

引用

8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022

作者： Phoon, K.K. Shuku, T. Archtecture and Sustainable Design/Information Systems Technology and Design Singapore University of Technology Singapore Graduate School of Environment and Life Science Okayama University Japan

ISBN: (纸本)9789811851841

Geotechnical lasso (Glasso), which is a data-drive site characterization (DDSC) method proposed by the authors, is advantageous to identification of abrupt changes in data such as soil layer boundaries because it is formulated with piecewise functions. In Glasso, target grounds are split into many cells or cuboids, and geotechnical parameters for all the cells/cuboids need to be identified. It requires massive computation time and memory space and has limitations for high degree-of-freedom (DoF) problems. This study newly proposed a Glasso based on continuous basis functions (BFs) to reduce DoF, i.e., to address the limitations of the existing Glasso. The applicability of the proposed Glasso with BFs is investigated through DDSC for virtual grounds. This study also shed a light on the advantages and disadvantages of two types of Glasso based on the comparison. © 2022 ISRERM Organizers. Published by Research Publishing, Singapore.

关键词： Degrees of freedom (mechanics)

来源：评论

学校读者我要写书评

暂无评论

Report for ISSMGE TC309/TC304/TC222 and ASCE Geo-Institute Risk Assessment and Management Committee Fourth Machine Learning in Geotechnics Dialogue on “Machine Learning Supremacy Projects”: 5 December 2023, Okayama Convention Center, Okayama, Japan

引用

Georisk 2024年第1期18卷 304-313页

作者： Leung, Andy Y.F. Phoon, Kok-Kwang Xiao, Te Shuku, Takayuki Ching, Jianye Department of Civil and Environmental Engineering The Hong Kong Polytechnic University Hong Kong Information Systems Technology and Design/Architecture and Sustainable Design Singapore University of Technology and Design Singapore Singapore Department of Civil and Environmental Engineering The Hong Kong University of Science and Technology Hong Kong Okayama University Okayama Japan Department of Civil Engineering National Taiwan University Taipei Taiwan

The ISSMGE TC309/TC304/TC222 and ASCE Geo-Institute Risk Assessment Management Committee Fourth Machine Learning (ML) in Geotechnics Dialogue was held at the Okayama Convention Center on 5 December 2023. The dialogue focused on “machine learning supremacy projects”, which involve the use of novel ML techniques to analyse emerging data, potentially bringing disruptive value to transform research and practice. The discussions centred around four real-world projects covering tunnelling, pile foundations, and geophysical investigation, to showcase the latest developments and implementations of ML, exchange views on the associated technical and non-technical challenges, and explore further developments to transform practice on a broader scale. These projects highlight the need for efficient ML techniques to analyse large volumes of real-time data for 3D heterogeneous soil/rock domains, and the benefits of standardising data formats among practitioners and researchers. Meanwhile, as data become valuable assets, there is a need to establish guidelines or protocols to protect data privacy and ensure data trustworthiness while creating incentives for owners to share data. Several recommendations include (1) creating benchmark examples using real data and Class A prediction (i.e. prediction made before the event) exercises from real projects to accelerate research in ML methods and (2) involving regulators in future dialogue and exploring the incorporation of data-driven approaches in design codes to complement traditional methods. © 2024 Informa UK Limited, trading as Taylor & Francis Group.

关键词： Big data data-centric geotechnics digital transformation emerging data machine learning supremacy projects

来源：评论

学校读者我要写书评

暂无评论

Geotechnical "facial Recognition" Challenge

引用

ASCE-ASME Journal of Risk and Uncertainty in Engineering systems, Part A: Civil Engineering 2025年第3期11卷

作者： Phoon, Kok-Kwang Cai, Yongmin Tang, Chong Information Systems Technology and Design/Architecture and Sustainable Design Singapore Univ. of Technology and Design Somapah Rd. 487372 Singapore Architecture and Sustainable Design Singapore Univ. of Technology and Design Somapah Rd. 487372 Singapore Faculty of Infrastructure Engineering Dalian Univ. of Technology Liaoning Dalian116024 China

The site recognition challenge is to quantify site uniqueness, directly or indirectly, so that sparse site-specific data can be supplemented by big indirect data (BID) to produce a quasi-site-specific model. Engineers appreciate that every site is unique to some extent, but they do not have the ability to quantify this uniqueness. In all likelihood, the difference between two sites depends on the feature of interest (stratification, spatial variability, properties, anomaly, etc.). All sites in a region share common general features as a result of their geologic origin, but they are different at a detailed level when investigated using site investigation tools. This is broadly similar to the facial recognition challenge. However, the training set for facial recognition (photographs) is much bigger and more complete than the available training set for the site recognition challenge. In fact, there is no 3D photograph of a ground volume for any feature of engineering interest. Nonetheless, for the case of soil properties, this paper shows that it is possible to construct a quasi-site-specific (or quasilocal) transformation model from site specific small data and big indirect data. The strategies include the hierarchical Bayesian method, tailored clustering method, outlier detection, dimension reduction, and dictionary learning. These Bayesian machine learning methods can be combined to produce inferences that are more accurate and less biased than those from the classical probabilistic multiple regression (PMR) method. The computational cost of PMR is the lowest, but its disadvantages are significant. One example is that PMR cannot handle multivariate, uncertain and unique, sparse, incomplete, and potentially corrupted (MUSIC) which is a common set of data attributes for real site data. This paper reviews the state-of-The-Art research in the site recognition challenge in the presence of MUSIC. Spatial variability and geologic uncertainty are not included in this review

关键词： Hierarchical clustering

来源：评论

学校读者我要写书评

暂无评论

Quantifying Form Diversity in Computational design Using a Phenotype-Based Strategy

SSRN

引用

SSRN 2024年

作者： Liang, Ling Ban Tunçer, Bige Architecture and Sustainable Design Singapore University of Technology and Design Singapore Information Systems in the Built Environment Eindhoven University of Technology Netherlands

Solution diversity refers to a spread of solutions that have different geometrical properties but address similar design goals. It is an arbitrary design quality which depends not just on quantity but also the ability of the solution to address design aspirations. Solution diversity often acts as an implicit design goal, and it influences one’s perception of the explored span of the design space. In computational design, diversity is measured using genotype-based or performance-based values. However, these strategies are relevant only to the current design scenario as the solution representation is specific to the current parametric model. This means that current computational methods are unable to match up to the speed of design exploration where multiple parametric models are tested. Therefore, this paper proposes a phenotype-based strategy that addresses the three main challenges in diversity quantification: 1) Identifying a relevant representation method for geometrical properties within a design space. 2) Identifying a representation method that is relevant across different design scenarios. 3) Ensuring that the identified representation method can be flexibly interpreted across scenarios. © 2024, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Performance Modeling of Discharging Process in a Thermochemical Fluid System Using Machine Learning Approaches 31

Performance Modeling of Discharging Process in a Thermochemi...

引用

31st International Workshop on Intelligent Computing in Engineering, EG-ICE 2024

作者： Mohammadi, Sahar Nabavi, Seyed Azad Buchholz, Reiner Geyer, Philipp Sustainable Building Systems Group Institute for Design and Construction Faculty of Architecture and Landscape Leibniz University Hannover Germany Institute for Energy Technology Technical University of Berlin Germany

Open absorption system is an energy efficient method for dehumidification/humidification. As the global demand for clean energy continues to rise, the significance of machine learning in forecasting energy systems will amplify. In this paper, we developed three machine learning algorithms to accurately forecast air output mixing ratio and temperature of a thermochemical fluid system and compare their performance. The variables used to forecast our responses are ambient air temperature (°C), inlet air mixing ratio (g/kg), the inlet desiccant temperature(°C), the air mass flow rate (kg/s) and desiccant mass flow rate (kg/s). The evaluation metrics shows Random Forest regressor greatly outperform multi-layer perceptron neural network, and is slightly better than Gradient boosting regressor. The Random Forest regressor stands out with a significant R2 of 0.999 for predicting both air output mixing ratio and temperature, accompanied by MAPE values of 0.241 and 0.097, for air output mixing ratio and temperature prediction, respectively. © 2024 Proceedings of the 31st International Workshop on Intelligent Computing in Engineering, EG-ICE 2024. All rights reserved.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

Analysis of Machine Learning (ML) and Internet of Things (IoT): A Scientometric Review 20

Analysis of Machine Learning (ML) and Internet of Things (Io...

引用

20th IEEE International Colloquium on Signal Processing and Its Applications, CSPA 2024

作者： Ajibade, Samuel-Soma M. Jasser, Muhammed Basheer Legaspino, Rosie Fe B. Gimeno, Anabella R. Ypanto, Leomar A. Baird, Jeem Clyde C. Adediran, Anthonia Oluwatosin Sunway University School Of Engineering And Technology Department Of Computing And Information Systems Selangor Malaysia Cebu Technological University Cebu Moalboal Philippines Federal University Of Uberlandia Faculty Of Architecture And Urban Design Uberlândia Brazil

ISBN: (纸本)9798350382310

The objective of this paper is to critically examine the publication's trends and research landscape on the Internet of Things and Machine Learning (Io T-ML) research through bibliometric analysis (BA). The Elsevier Scopus database was selected to identify, screen, and analyze the published documents on IoT-ML. The publication trends and major stakeholders on the topic were subsequently. BA was performed to examine the co-authorships, keyword occurrences, and citations using VOSviewer software. Lastly, a systematic literature review was performed to summarize the scientific and technological advancements. Results revealed that the publications on IoT-ML largely comprise articles, conference proceedings, and reviews, whereas the publication trends showed an increase from 2 to 105 between 2015 and 2021. The top researchers on IoT-ML are S. Baskar, M. Guizani, and A. Souri, whereas the top institution is the Vellore Institute of technology (India). However, the top country and research funder are;India, and National Natural Science Foundation (China), respectively. BA showed that the research landscape on IoT-ML comprises numerous networks of co-authors and countries working on related themes. The most impactful publication is Mahdavinejad et al. (2018) which has 479 citations, whereas the most prestigious journal is the IEEE Io T journal with 9 published documents in Scopus. Keyword analysis revealed IoT-ML has potential applications in health care, medicine, computer security and data analytics etc. Overall, IoT-ML has significant socio-economic and research impacts, and as such, the field will witness geometric techno-scientific growth and development in the future, particularly in the areas of network security and big data analytics. © 2024 IEEE.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：