首页|Reports Outline Machine Learning Findings from University of Birmingham (Optimis ing Synthetic Datasets for Machine Learningbased Prediction of Building Damage Due To Tunnelling)

Reports Outline Machine Learning Findings from University of Birmingham (Optimis ing Synthetic Datasets for Machine Learningbased Prediction of Building Damage Due To Tunnelling)

扫码查看
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Current study results on Machine Learn ing have been published. According to news reporting out of Birmingham, United K ingdom, by NewsRx editors, research stated, "Assessment of tunnelling-induced bu ilding damage is a complex Soil-Structure Interaction (SSI) probelm, influenced by numerous geometric and material parameters of both the soil and structures, a nd is characterised by strong non-linear behaviour. Currently, there is a trend towards developing data-driven models using Machine Learning (ML) to capture thi s complex behaviour." Our news journalists obtained a quote from the research from the University of B irmingham, "Given the scarcity of real data, which typically comes from specific case studies, many researchers have turned to creating extensive synthetic data sets via sophisticated and validated numerical models like Finite Element Method (FEM). However, the development of these datasets and the training of advanced ML algorithms present significant challenges. poses significant challenges. Reli ance solely on parameter domains and ranges derived from case studies can lead t o imbalanced data distributions and subsequently poor performance of models in l ess populated regions. In this paper, we introduce a strategy for designing opti mal high-confidence datasets through an iterative procedure. This process begins with a systematic literature review to determine the importance of parameters, their ranges, and dependencies as they pertain to building damage induced by SSI . Starting with several hundred FEM simulations, we generate an initial dataset and assess its quality and impact through Sensitivity Analysis (SA) studies, sta tistical modelling, and re-sampling in statistically significant regions. This e valuation allows us to refine the model's input space, seeking scenarios that mi tigate output distribution imbalances. The procedure is repeated until the datas ets achieve a satisfactory balance for training metamodels, minimising bias effe ctively. Our findings highlight the success of this approach in identifying an o ptimal and feasible input space that significantly reduces imbalanced distributi ons of output features."

BirminghamUnited KingdomEuropeCybo rgsEmerging TechnologiesMachine LearningUniversity of Birmingham

2024

Robotics & Machine Learning Daily News

Robotics & Machine Learning Daily News

ISSN:
年,卷(期):2024.(Oct.4)