The Experts below are selected from a list of 531975 Experts worldwide ranked by ideXlab platform
Steven Euijong Whang - One of the best experts on this subject based on the ideXlab platform.
-
automated data slicing for Model Validation a big data ai integration approach
IEEE Transactions on Knowledge and Data Engineering, 2020Co-Authors: Yeounoh Chung, Tim Kraska, Neoklis Polyzotis, Ki Hyun Tae, Steven Euijong WhangAbstract:As machine learning systems become democratized, it becomes increasingly important to help users easily debug their Models. However, current data tools are still primitive when it comes to helping users trace Model performance problems all the way to the data. We focus on the particular problem of slicing data to identify subsets of the Validation data where the Model performs poorly. This is an important problem in Model Validation because the overall Model performance can fail to reflect that of the smaller subsets, and slicing allows users to analyze the Model performance on a more granular-level. Unlike general techniques (e.g., clustering) that can find arbitrary slices, our goal is to find interpretable slices (which are easier to take action compared to arbitrary subsets) that are problematic and large. We propose $\mathsf{Slice Finder}$ SliceFinder , which is an interactive framework for identifying such slices using statistical techniques. Applications include diagnosing Model fairness and fraud detection, where identifying slices that are interpretable to humans is crucial. This research is part of a larger trend of Big data and Artificial Intelligence (AI) integration and opens many opportunities for new research.
-
slice finder automated data slicing for Model Validation
International Conference on Data Engineering, 2019Co-Authors: Yeounoh Chung, Tim Kraska, Neoklis Polyzotis, Ki Hyun Tae, Steven Euijong WhangAbstract:As machine learning (ML) systems become democratized, it becomes increasingly important to help users easily debug their Models. However, current data tools are still primitive when it comes to helping users trace Model performance problems all the way to the data. We focus on the particular problem of slicing data to identify subsets of the Validation data where the Model performs poorly. This is an important problem in Model Validation because the overall Model performance can fail to reflect that of the smaller subsets, and slicing allows users to analyze the Model performance on a more granular-level. Unlike general techniques (e.g., clustering) that can find arbitrary slices, our goal is to find interpretable slices (which are easier to take action compared to arbitrary subsets) that are large and problematic. We propose Slice Finder, which is an interactive framework for identifying such slices using statistical techniques. Applications include diagnosing Model fairness and fraud detection, where identifying slices that are interpretable to humans is crucial.
-
automated data slicing for Model Validation a big data ai integration approach
arXiv: Databases, 2018Co-Authors: Yeounoh Chung, Tim Kraska, Neoklis Polyzotis, Ki Hyun Tae, Steven Euijong WhangAbstract:As machine learning systems become democratized, it becomes increasingly important to help users easily debug their Models. However, current data tools are still primitive when it comes to helping users trace Model performance problems all the way to the data. We focus on the particular problem of slicing data to identify subsets of the Validation data where the Model performs poorly. This is an important problem in Model Validation because the overall Model performance can fail to reflect that of the smaller subsets, and slicing allows users to analyze the Model performance on a more granular-level. Unlike general techniques (e.g., clustering) that can find arbitrary slices, our goal is to find interpretable slices (which are easier to take action compared to arbitrary subsets) that are problematic and large. We propose Slice Finder, which is an interactive framework for identifying such slices using statistical techniques. Applications include diagnosing Model fairness and fraud detection, where identifying slices that are interpretable to humans is crucial. This research is part of a larger trend of Big data and Artificial Intelligence (AI) integration and opens many opportunities for new research.
Haifeng Gong - One of the best experts on this subject based on the ideXlab platform.
-
Variable thickness scroll compressor performance analysis—Part II: Dynamic Modeling and Model Validation
2017Co-Authors: Legros Arnaud, Hongsheng Zhang, Haifeng GongAbstract:In order to investigate the performance of variable thickness scroll compressor, a detail mathematical Modeling based on energy and mass balances is established in this two-part. In part II, dynamic Modeling and Model Validation are developed. Temperature, pressure, mass flow of working chambers, friction loss power of moving parts, efficiency, and shaft power are investigated by solving the mathematical Modeling. The experimental rig for variable thickness scroll compressor based on involute of circle, high order curve and arc is set up. From the comparison of the simulated and measured data, it can be seen that the compressor Model predicts the mass flow, discharge temperature, and shaft power very well. So the proposed mathematical Modeling can accurately describe all the suction, compression, and discharge processes for variable thickness scroll compressor.Peer reviewe
-
Variable thickness scroll compressor performance analysis—Part II: Dynamic Modeling and Model Validation
'SAGE Publications', 2017Co-Authors: Bin Peng, Lemort Vincent, Legros Arnaud, Hongsheng Zhang, Haifeng GongAbstract:peer reviewedaudience: researcher, professionalIn order to investigate the performance of variable thickness scroll compressor, a detail mathematical Modeling based on energy and mass balances is established in this two-part. In part II, dynamic Modeling and Model Validation are developed. Temperature, pressure, mass flow of working chambers, friction loss power of moving parts, efficiency, and shaft power are investigated by solving the mathematical Modeling. The experimental rig for variable thickness scroll compressor based on involute of circle, high order curve and arc is set up. From the comparison of the simulated and measured data, it can be seen that the compressor Model predicts the mass flow, discharge temperature, and shaft power very well. So the proposed mathematical Modeling can accurately describe all the suction, compression, and discharge processes for variable thickness scroll compressor
Yeounoh Chung - One of the best experts on this subject based on the ideXlab platform.
-
automated data slicing for Model Validation a big data ai integration approach
IEEE Transactions on Knowledge and Data Engineering, 2020Co-Authors: Yeounoh Chung, Tim Kraska, Neoklis Polyzotis, Ki Hyun Tae, Steven Euijong WhangAbstract:As machine learning systems become democratized, it becomes increasingly important to help users easily debug their Models. However, current data tools are still primitive when it comes to helping users trace Model performance problems all the way to the data. We focus on the particular problem of slicing data to identify subsets of the Validation data where the Model performs poorly. This is an important problem in Model Validation because the overall Model performance can fail to reflect that of the smaller subsets, and slicing allows users to analyze the Model performance on a more granular-level. Unlike general techniques (e.g., clustering) that can find arbitrary slices, our goal is to find interpretable slices (which are easier to take action compared to arbitrary subsets) that are problematic and large. We propose $\mathsf{Slice Finder}$ SliceFinder , which is an interactive framework for identifying such slices using statistical techniques. Applications include diagnosing Model fairness and fraud detection, where identifying slices that are interpretable to humans is crucial. This research is part of a larger trend of Big data and Artificial Intelligence (AI) integration and opens many opportunities for new research.
-
slice finder automated data slicing for Model Validation
International Conference on Data Engineering, 2019Co-Authors: Yeounoh Chung, Tim Kraska, Neoklis Polyzotis, Ki Hyun Tae, Steven Euijong WhangAbstract:As machine learning (ML) systems become democratized, it becomes increasingly important to help users easily debug their Models. However, current data tools are still primitive when it comes to helping users trace Model performance problems all the way to the data. We focus on the particular problem of slicing data to identify subsets of the Validation data where the Model performs poorly. This is an important problem in Model Validation because the overall Model performance can fail to reflect that of the smaller subsets, and slicing allows users to analyze the Model performance on a more granular-level. Unlike general techniques (e.g., clustering) that can find arbitrary slices, our goal is to find interpretable slices (which are easier to take action compared to arbitrary subsets) that are large and problematic. We propose Slice Finder, which is an interactive framework for identifying such slices using statistical techniques. Applications include diagnosing Model fairness and fraud detection, where identifying slices that are interpretable to humans is crucial.
-
automated data slicing for Model Validation a big data ai integration approach
arXiv: Databases, 2018Co-Authors: Yeounoh Chung, Tim Kraska, Neoklis Polyzotis, Ki Hyun Tae, Steven Euijong WhangAbstract:As machine learning systems become democratized, it becomes increasingly important to help users easily debug their Models. However, current data tools are still primitive when it comes to helping users trace Model performance problems all the way to the data. We focus on the particular problem of slicing data to identify subsets of the Validation data where the Model performs poorly. This is an important problem in Model Validation because the overall Model performance can fail to reflect that of the smaller subsets, and slicing allows users to analyze the Model performance on a more granular-level. Unlike general techniques (e.g., clustering) that can find arbitrary slices, our goal is to find interpretable slices (which are easier to take action compared to arbitrary subsets) that are problematic and large. We propose Slice Finder, which is an interactive framework for identifying such slices using statistical techniques. Applications include diagnosing Model fairness and fraud detection, where identifying slices that are interpretable to humans is crucial. This research is part of a larger trend of Big data and Artificial Intelligence (AI) integration and opens many opportunities for new research.
Hermann G Matthies - One of the best experts on this subject based on the ideXlab platform.
-
epistemic uncertainty based Model Validation via interval propagation and parameter calibration
Computer Methods in Applied Mechanics and Engineering, 2018Co-Authors: Chong Wang, Hermann G MatthiesAbstract:Abstract The Model Validation with respect to epistemic uncertainty , where only a small amount of experimental data is available, is a challenging problem in practical engineering. Interval theory is a useful tool to deal with such epistemic uncertainty, and this paper aims to construct an interval theory-based analysis framework for Model Validation. Using the statistical moments of experimental observations, an unbiased estimation method is firstly presented to quantify the interval bounds of system uncertainties. In the subsequent process of predicting uncertain responses, the Legendre polynomial chaos expansion is introduced as the surrogate Model , which can greatly improve the computational efficiency of uncertainty propagation . Then the concept of interval fitting degree is proposed to establish a new quantitative Validation metric, which can accurately characterize the agreement between the computational response interval and experimental response interval. Meanwhile, an operation of interval parameter calibration is executed in the form of optimization to improve the prediction accuracy of computational Model. Finally, the Sandia thermal challenge problem is utilized to verify the feasibility of presented Model Validation method in engineering application.
-
evidence theory based Model Validation method for heat transfer system with epistemic uncertainty
International Journal of Thermal Sciences, 2018Co-Authors: Chong Wang, Hermann G MatthiesAbstract:Abstract In numerical heat transfer, the Model Validation problem with respect to epistemic uncertainty, where only a small amount of experimental information is available, has been recognized as a challenging issue. To overcome the drawback of traditional probabilistic methods in dealing with limited data, this paper proposes a novel Model Validation approach by using evidence theory. First, the evidence variables are adopted to characterize the uncertain input parameters, where the focal elements are expressed as mutually connected intervals with basic probability assignment (BPA). In the subsequent process of predicting response focal elements, an interval collocation analysis method with small computational cost is presented. By combining the response BPAs in both experimental measurements and numerical predictions, a new parameter calibration method is then developed to further improve the accuracy of computational Model. Meanwhile, an evidence-theory-based Model Validation metric is defined to test the Model credibility. Eventually, the famous Sandia thermal challenge problem is utilized to verify the feasibility of presented Model Validation method in engineering application.
Wei Jiang - One of the best experts on this subject based on the ideXlab platform.
-
parameter sensitivity examination and discussion of pem fuel cell simulation Model Validation part i current status of Modeling research and Model development
Journal of Power Sources, 2006Co-Authors: Wenquan Tao, C H Min, Xiaobin Liu, B H Yin, Wei JiangAbstract:Mathematical Modeling plays an important role in fuel cell design. A comprehensive review of the mathematical Modeling of proton exchange membrane fuel cells is first conducted. It is found that the results computed by different Models in the literature often agree well with the experimental data. This stimulates the present authors to carry out a comprehensive parameter sensitivity examination. In this first paper a three-dimensional, two-phase and non-isothermal Model is developed, and numerical simulations for a basic case is performed, the results of which are regarded as the reference for further sensitivity examination. All the parameters needed for the simulation are provided in detail. In the companion paper (Part II), the results of the parameter sensitivity analyses and discussion of Model Validation are provided in detail.
