Active Learning of Post-earthquake Structural Damage with Co-optimal Information Gain and Reconnaissance Cost

2021 ◽  
pp. 9-16
Author(s):  
Mohamadreza Sheibani ◽  
Ge Ou
Keyword(s):  
Active Learning ◽  
Structural Damage ◽  
Information Gain ◽  
Optimal Information

scholarly journals Bayesian3 Active Learning for the Gaussian Process Emulator Using Information Theory

Entropy ◽  
2020 ◽  
Vol 22 (8) ◽  
pp. 890
Author(s):  
Sergey Oladyshkin ◽  
Farid Mohammadi ◽  
Ilja Kroeker ◽  
Wolfgang Nowak
Keyword(s):  
Bayesian Inference ◽  
Active Learning ◽  
Gaussian Process ◽  
Relative Entropy ◽  
Bayesian Model ◽  
Information Gain ◽  
Measurement Data ◽  
Superior Performance ◽  
Evidence Based ◽  
Model Evidence

Gaussian process emulators (GPE) are a machine learning approach that replicates computational demanding models using training runs of that model. Constructing such a surrogate is very challenging and, in the context of Bayesian inference, the training runs should be well invested. The current paper offers a fully Bayesian view on GPEs for Bayesian inference accompanied by Bayesian active learning (BAL). We introduce three BAL strategies that adaptively identify training sets for the GPE using information-theoretic arguments. The first strategy relies on Bayesian model evidence that indicates the GPE’s quality of matching the measurement data, the second strategy is based on relative entropy that indicates the relative information gain for the GPE, and the third is founded on information entropy that indicates the missing information in the GPE. We illustrate the performance of our three strategies using analytical- and carbon-dioxide benchmarks. The paper shows evidence of convergence against a reference solution and demonstrates quantification of post-calibration uncertainty by comparing the introduced three strategies. We conclude that Bayesian model evidence-based and relative entropy-based strategies outperform the entropy-based strategy because the latter can be misleading during the BAL. The relative entropy-based strategy demonstrates superior performance to the Bayesian model evidence-based strategy.

scholarly journals A unifying computational framework for teaching and active learning

2018 ◽  
Author(s):  
Scott Cheng-Hsin Yang ◽  
Wai Keen Vong ◽  
Yue Yu ◽  
Patrick Shafto
Keyword(s):  
Theory Of Mind ◽  
Active Learning ◽  
Information Gain ◽  
Learning Problems ◽  
Teaching Model ◽  
Simulation Experiments ◽  
Computational Framework ◽  
Learning And Teaching ◽  
Core Cognition ◽  
Present Simulation

Traditionally learning has been modeled as passively obtaining information or actively exploring the environment. Recent research has introduced models of learning from teachers that involve reasoning about why they have selected particular evidence. We introduce a computational framework that takes a critical step toward unifying active learning and teaching by recognizing that meta reasoning underlying reasoning about others can be applied to reasoning about oneself. The resulting Self-Teaching model captures much of the behavior of information-gain-based active learning with elements of hypothesis-testing-based active learning and can thus be considered as a formalization of active learning within the broader teaching framework. We present simulation experiments that characterize the behavior of the model within three simple and well-investigated learning problems. We conclude by discussing such theory-of-mind-based learning in the context of core cognition and cognitive development.

Bayesian Active Learning of Neural Firing Rate Maps with Transformed Gaussian Process Priors

2014 ◽  
Vol 26 (8) ◽  
pp. 1519-1541 ◽  
Author(s):  
Mijung Park ◽  
J. Patrick Weller ◽  
Gregory D. Horwitz ◽  
Jonathan W. Pillow
Keyword(s):  
Active Learning ◽  
Gaussian Process ◽  
Firing Rate ◽  
Utility Function ◽  
Information Gain ◽  
Tuning Curve ◽  
Strong Dependence ◽  
Learning Performance ◽  
Stimulus Selection ◽  
Neural Firing

A firing rate map, also known as a tuning curve, describes the nonlinear relationship between a neuron's spike rate and a low-dimensional stimulus (e.g., orientation, head direction, contrast, color). Here we investigate Bayesian active learning methods for estimating firing rate maps in closed-loop neurophysiology experiments. These methods can accelerate the characterization of such maps through the intelligent, adaptive selection of stimuli. Specifically, we explore the manner in which the prior and utility function used in Bayesian active learning affect stimulus selection and performance. Our approach relies on a flexible model that involves a nonlinearly transformed gaussian process (GP) prior over maps and conditionally Poisson spiking. We show that infomax learning, which selects stimuli to maximize the information gain about the firing rate map, exhibits strong dependence on the seemingly innocuous choice of nonlinear transformation function. We derive an alternate utility function that selects stimuli to minimize the average posterior variance of the firing rate map and analyze the surprising relationship between prior parameterization, stimulus selection, and active learning performance in GP-Poisson models. We apply these methods to color tuning measurements of neurons in macaque primary visual cortex.

Active learning data selection for adaptive online structural damage estimation

2010 ◽  
Author(s):  
Debejyo Chakraborty ◽  
Narayan Kovvali ◽  
Antonia Papandreou-Suppappola ◽  
Aditi Chattopadhyay
Keyword(s):  
Active Learning ◽  
Structural Damage ◽  
Data Selection ◽  
Damage Estimation ◽  
Selection For ◽  
Learning Data

scholarly journals Active Learning Performance in Labeling Radiology Images Is 90% Effective

2021 ◽  
Vol 1 ◽  
Author(s):  
Patrick Bangert ◽  
Hankyu Moon ◽  
Jae Oh Woo ◽  
Sima Didari ◽  
Heng Hao
Keyword(s):  
Artificial Intelligence ◽  
Active Learning ◽  
Information Content ◽  
Information Gain ◽  
Real Life ◽  
Learning Performance ◽  
Image Labeling ◽  
Human Effort ◽  
Polygonal Shape ◽  
Entire Dataset

To train artificial intelligence (AI) systems on radiology images, an image labeling step is necessary. Labeling for radiology images usually involves a human radiologist manually drawing a (polygonal) shape onto the image and attaching a word to it. As datasets are typically large, this task is repetitive, time-consuming, error-prone, and expensive. The AI methodology of active learning (AL) can assist human labelers by continuously sorting the unlabeled images in order of information gain and thus getting the labeler always to label the most informative image next. We find that after about 10%, depending on the dataset, of the images in a realistic dataset are labeled, virtually all the information content has been learnt and the remaining images can be automatically labeled. These images can then be checked by the radiologist, which is far easier and faster to do. In this way, the entire dataset is labeled with much less human effort. We introduce AL in detail and expose the effectiveness using three real-life datasets. We contribute five distinct elements to the standard AL workflow creating an advanced methodology.

Evaluation of single-electron movies of glutamine synthetase molecules

1983 ◽  
Vol 41 ◽  
pp. 280-281
Author(s):  
W. Kunath ◽  
E. Zeitler ◽  
M. Kessel
Keyword(s):  
Electron Microscope ◽  
Glutamine Synthetase ◽  
Electron Irradiation ◽  
Structural Damage ◽  
Structural Changes ◽  
Single Electron ◽  
Continuous Series ◽  
Digital Recording ◽  
Recording Device ◽  
Low Exposure

The features of digital recording of a continuous series (movie) of singleelectron TV frames are reported. The technique is used to investigate structural changes in negatively stained glutamine synthetase molecules (GS) during electron irradiation and, as an ultimate goal, to look for the molecules' “undamaged” structure, say, after a 1 e/Å2 dose.The TV frame of fig. la shows an image of 5 glutamine synthetase molecules exposed to 1/150 e/Å2. Every single electron is recorded as a unit signal in a 256 ×256 field. The extremely low exposure of a single TV frame as dictated by the single-electron recording device including the electron microscope requires accumulation of 150 TV frames into one frame (fig. lb) thus achieving a reasonable compromise between the conflicting aspects of exposure time per frame of 3 sec. vs. object drift of less than 1 Å, and exposure per frame of 1 e/Å2 vs. rate of structural damage.

Temperature Dependence of the Critical Electron Dose for Fatty Acid Monolayers

1982 ◽  
Vol 40 ◽  
pp. 78-79
Author(s):  
Kenneth H. Downing ◽  
Robert M. Glaeser
Keyword(s):  
Electron Beam ◽  
Fatty Acid ◽  
Inelastic Scattering ◽  
Temperature Dependence ◽  
Structural Damage ◽  
Direct Result ◽  
Second Step ◽  
Strong Temperature Dependence ◽  
Electron Dose ◽  
Covalent Bonds

The structural damage of molecules irradiated by electrons is generally considered to occur in two steps. The direct result of inelastic scattering events is the disruption of covalent bonds. Following changes in bond structure, movement of the constituent atoms produces permanent distortions of the molecules. Since at least the second step should show a strong temperature dependence, it was to be expected that cooling a specimen should extend its lifetime in the electron beam. This result has been found in a large number of experiments, but the degree to which cooling the specimen enhances its resistance to radiation damage has been found to vary widely with specimen types.

Modification of the Electron Microscope for Investigation of Fully Hydrated Biological Specimens

1969 ◽  
Vol 27 ◽  
pp. 418-419 ◽  
Author(s):  
R. C. Moretz ◽  
D. F. Parsons
Keyword(s):  
Electron Beam ◽  
Electron Microscope ◽  
Structural Damage ◽  
Lower Percentage ◽  
Vacuum Drying ◽  
Total Removal ◽  
Total Absence ◽  
Biological Studies ◽  
Electron Microscopes ◽  
Beam Damage

Short lifetime or total absence of electron diffraction of ordered biological specimens is an indication that the specimen undergoes extensive molecular structural damage in the electron microscope. The specimen damage is due to the interaction of the electron beam (40-100 kV) with the specimen and the total removal of water from the structure by vacuum drying. The lower percentage of inelastic scattering at 1 MeV makes it possible to minimize the beam damage to the specimen. The elimination of vacuum drying by modification of the electron microscope is expected to allow more meaningful investigations of biological specimens at 100 kV until 1 MeV electron microscopes become more readily available. One modification, two-film microchambers, has been explored for both biological and non-biological studies.

Sign in / Sign up

Export Citation Format

Share Document