Filters








5,106 Hits in 6.2 sec

Efficient Parameter Importance Analysis via Ablation with Surrogates

Andre Biedenkapp, Marius Lindauer, Katharina Eggensperger, Frank Hutter, Chris Fawcett, Holger Hoos
2017 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Here, we show how the running time cost of ablation analysis, a well-known general-purpose approach for assessing parameter importance, can be reduced substantially by using regression models of algorithm  ...  While impressive performance gains are often achieved in this manner, additional, potentially costly parameter importance analysis is required to gain insights into what parameter changes are most responsible  ...  Efficient Ablation via Surrogates As previously mentioned, ablation analysis is an expensive process due to the algorithm runs that have to be performed to gather enough empirical evidence to compare parameter  ... 
doi:10.1609/aaai.v31i1.10657 fatcat:4vbl543eovhczfo5cfmzqfsjbi

Facilitating Database Tuning with Hyper-Parameter Optimization: A Comprehensive Experimental Evaluation [article]

Xinyi Zhang, Zhuo Chang, Yang Li, Hong Wu, Jian Tan, Feifei Li, Bin Cui
2022 arXiv   pre-print
Beyond the comprehensive evaluations, we offer an efficient and unified database configuration tuning benchmark via surrogates that reduces the evaluation cost to a minimum, allowing for extensive runs  ...  and analysis of new techniques.  ...  The performance evaluation is approximated via surrogate for efficiency reasons. The order that the feature is changed is the importance rank from the ablation analysis.  ... 
arXiv:2110.12654v4 fatcat:rfzh4rpp6relxfcmlu3qbtsg2m

Trajectory-Based Off-Policy Deep Reinforcement Learning [article]

Andreas Doerr, Michael Volpp, Marc Toussaint, Sebastian Trimpe, Christian Daniel
2019 arXiv   pre-print
Incorporation of previous rollouts via importance sampling greatly improves data-efficiency, whilst stochastic optimization schemes facilitate the escape from local optima.  ...  This work addresses these weaknesses by combining recent improvements in the reuse of off-policy data and exploration in parameter space with deterministic behavioral policies.  ...  In the hyper-parameter tuning phase, experiments with TRPO and PPO have been conducted based on smaller batchsizes, but due to the lack of data-efficient incorporation of off-policy data, no faster and  ... 
arXiv:1905.05710v1 fatcat:6po2azo7yndsrjmh4ewcdnfmum

Catheter Ablation of Atrial Fibrillation: A Review of the Current Status and Future Directions

Daniel Melby
2017 Journal of Innovations in Cardiac Rhythm Management  
Over the last 20 years, the frequency of use of catheter ablation to treat AF has grown, commensurate with the rise in arrhythmia burden and via a number of technical advancements.  ...  These developments can be divided into new techniques for myocardial ablation, improvements in the understanding of AF trigger mechanisms, and advancements in atrial mapping.  ...  As a result, ablation time was often the only parameter under direct and measurable control of the electrophysiologist.  ... 
doi:10.19102/icrm.2017.081101 pmid:32477760 pmcid:PMC7252758 fatcat:e54bnlv5pnc7lpkryyvyhlmsau

Black-box Gradient Attack on Graph Neural Networks: Deeper Insights in Graph-based Attack and Defense [article]

Haoxi Zhan, Xiaobing Pei
2021 arXiv   pre-print
As a result, a simple yet useful strategy to defense against Mettack is to train the GNN with the validation set.  ...  While being able to successfully decrease the performance of GNNs, most existing attacking algorithms require access to either the model parameters or the training data, which is not practical in the real  ...  Ablation Study and Parameter Analysis To answer the research question RQ3, we conducted ablation studies and parameter analysis. For ablation study, we created two variant of our method.  ... 
arXiv:2104.15061v2 fatcat:a7odz5p5mjb73pzb6urkrfajcu

Are Gradients on Graph Structure Reliable in Gray-box Attacks? [article]

Zihan Liu, Yun Luo, Lirong Wu, Siyuan Li, Zicheng Liu, Stan Z. Li
2022 arXiv   pre-print
We propose edge discrete sampling to select the edge perturbations associated with hierarchical candidate selection to ensure computational efficiency.  ...  Previous gray-box attackers employ gradients from the surrogate model to locate the vulnerable edges to perturb the graph structure.  ...  The random initialization of parameters leads to variance in surrogate model training, which affects the structural gradients via back-propagation.  ... 
arXiv:2208.05514v1 fatcat:wpp46ix3azgt3o4ux2v3qgwude

Bayesian Optimisation for Sequential Experimental Design with Applications in Additive Manufacturing [article]

Mimi Zhang, Andrew Parnell, Dermot Brabazon, Alessio Benavoli
2021 arXiv   pre-print
This article is aimed at readers with some understanding of Bayesian methods, but not necessarily with knowledge of additive manufacturing; the software performance overview and implementation instructions  ...  Matern kernels are parameterized by a smoothness parameter ν > 0, and samples from a GP with a higher ν value are more smoother.  ...  We want to quickly determine an optimal parameter setting such that the productivity (ablated mass per ablation time interval) is high and the average particle size is small.  ... 
arXiv:2107.12809v3 fatcat:fvw3dmx2s5azpje5cs6dmhqkni

Learning Composable Energy Surrogates for PDE Order Reduction [article]

Alex Beatson, Jordan T. Ash, Geoffrey Roeder, Tianju Xue, Ryan P. Adams
2020 arXiv   pre-print
Costly ground-truth simulation of the full structure is avoided, as training data are generated by performing finite element analysis with individual components.  ...  Meta-materials are an important emerging class of engineered materials in which complex macroscopic behaviour--whether electromagnetic, thermal, or mechanical--arises from modular substructure.  ...  The effects of these design choices are quantified via an ablation study in the appendix. Reduced-basis parameterization. We require a vector representation for the functionũ.  ... 
arXiv:2005.06549v2 fatcat:tqzgdkdj7vdxrgliglpwqqilhq

Short-Time Estimation of Fractionation in Atrial Fibrillation with Coarse-Grained Correlation Dimension for Mapping the Atrial Substrate

Aikaterini Vraka, Fernando Hornero, Vicente Bertomeu-González, Joaquín Osca, Raúl Alcaraz, José J. Rieta
2020 Entropy  
Atrial fibrillation (AF) is currently the most common cardiac arrhythmia, with catheter ablation (CA) of the pulmonary veins (PV) being its first line therapy.  ...  The receiver operating characteristics (ROC) analysis for highly fractionated EGMs, showed 100% specificity and sensitivity in Group 1 and 87.5% specificity and 93.6% sensitivity in Group 2.  ...  It is important, therefore, to choose r cg as a trade-off between these two parameters.  ... 
doi:10.3390/e22020232 pmid:33286006 fatcat:ewy4tvv2hzat5jbumugb2t4kee

Quantifying the Uncertainty in Model Parameters Using Gaussian Process-Based Markov Chain Monte Carlo: An Application to Cardiac Electrophysiological Models [chapter]

Jwala Dhamala, John L. Sapp, Milan Horacek, Linwei Wang
2017 Lecture Notes in Computer Science  
A common solution is to replace the simulation model with a computationally-efficient surrogate for a faster sampling.  ...  Estimation of patient-specific model parameters is important for personalized modeling, although sparse and noisy clinical data can introduce significant uncertainty in the estimated parameter values.  ...  Below we describe the presented method that accelerates MCMC sampling of (3) via the use of an efficient GP surrogate in the modification of proposal distributions.  ... 
doi:10.1007/978-3-319-59050-9_18 fatcat:v2shjwc2tvab7kon74sbn6wz7y

Exploring Sensitivity of ICF Outputs to Design Parameters in Experiments Using Machine Learning [article]

Julia B. Nakhleh, M. Giselle Fernández-Godino, Michael J. Grosskopf, Brandon M. Wilson, John Kline, Gowri Srinivasan
2020 arXiv   pre-print
We present work using random forest (RF) regression for prediction of yield, velocity, and other experimental outcomes given a suite of design parameters, along with an assessment of important relationships  ...  In this paper, we leverage developments in machine learning (ML) and methods for ML feature importance/sensitivity analysis to identify complex relationships in ways that are difficult to process using  ...  The velocity of a DT layered implosion is inferred via a surrogate convergent ablator that uses X-ray radiography to observe capsule radius as a function of time.  ... 
arXiv:2010.04254v1 fatcat:ff2sno3u4zf7los55izw6f3boi

PiRank: Scalable Learning To Rank via Differentiable Sorting [article]

Robin Swezey, Aditya Grover, Bruno Charron, Stefano Ermon
2021 arXiv   pre-print
A key challenge with machine learning approaches for ranking is the gap between the performance metrics of interest and the surrogate loss functions that can be optimized with gradient-based methods.  ...  This gap arises because ranking metrics typically involve a sorting operation which is not differentiable w.r.t. the model parameters.  ...  The PiRank surrogate learning objective can be optimized via two gradient-based techniques in practice.  ... 
arXiv:2012.06731v2 fatcat:amtphimdijcodefysa67wxv2mu

Utilization of MOSES Modulated Pulse Mode Results in Improved Efficiency in Holmium:YAG laser Ablation of the Prostate

Bristol B. Whiles, Austin J. Martin, Andrew Brevik, Raphael V. Carrera, Jeffrey A. Thompson, Wilson R. Molina, Kerri L. Thurmon
2021 Urology  
However, HoLAP with MOSES had significantly higher ablation efficiency (0.59±0.24 g/min without vs. 0.86 0.5 g/min with MOSES; p=0.01).  ...  Ablation time was similar at 49.6±26.1 minutes without and 40.7±41.2 minutes with MOSES (p=0.38).  ...  Because 20 procedures had missing ablation times, we calculated a surrogate for ablation efficiency utilizing total OR time (grams/total OR time), which also showed improved efficiency with use of MOSES  ... 
doi:10.1016/j.urology.2020.12.024 pmid:33412223 fatcat:vtqti3hcjbg2blyqhsxzsxhxwq

Germline replacement by blastula cell transplantation in the fish medaka

Mingyou Li, Ni Hong, Hongyan Xu, Jianxing Song, Yunhan Hong
2016 Scientific Reports  
surrogate production 2,8-10 .  ...  It has remained largely unknown whether BCT is able to achieve GR at a high efficiency. Here we report efficient GR by BCT into blastula embryos in the fish medaka (Oryzias latipes).  ...  Our finding that BCT can achieve a high efficiency of fertility restoration and GR makes medaka an ideal model organism for the experimental analysis of these biological parameters towards BCT-mediated  ... 
doi:10.1038/srep29658 pmid:27406328 pmcid:PMC4942801 fatcat:gal6xmwb6zehthcx2e4umkc6sy

Biophysical Parameters During Radiofrequency Catheter Ablation of Scar-Mediated Ventricular Tachycardia: Epicardial and Endocardial Applications via Manual and Magnetic Navigation

TARA BOURKE, ERIC BUCH, NILESH MATHURIA, YOAV MICHOWITZ, RICKY YU, RAVI MANDAPATI, KALYANAM SHIVKUMAR, RODERICK TUNG
2014 Cardiovascular Electrophysiology  
with open-irrigation.  ...  There is a paucity of data on biophysical parameters during radiofrequency ablation of scarmediated ventricular tachycardia (VT).  ...  As interventional MR-guided ablation continues to evolve, studies correlating these biophysical surrogate parameters with transmural lesions will provide additional insight.  ... 
doi:10.1111/jce.12477 pmid:24946895 pmcid:PMC4282185 fatcat:h3sljtfpa5axvd76wanzbl7cje
« Previous Showing results 1 — 15 out of 5,106 results