Stochastic adaptive learning rate in an identification method: An approach for on-line drilling processes monitoring

Ba, Amadou; Hbaieb, S.; Mechbal, Nazih; Vergé, Michel

doi:10.1109/acc.2009.5159945

Cited by 3 publications

(3 citation statements)

References 9 publications

(14 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We trained Unet for N e = 100 epochs epochs with Adam [8], with a constant learning rate of 3 × 10 −4 , a weight decay parameter of 0.1 and a batch size of 12 on a single Tesla V100. The learning rate was unchanged during training to ensure parameter space exploration, using an analoguous reasoning to the one provided in [15].…”

Section: Setting Of Model Parametersmentioning

confidence: 99%

Ensembling Unets, sparse representation and low dimensional visualization for rare chromosomal aberration detection in light microscopy images

Deschemps,

Grégoire,

Martinez

et al. 2023

Preprint

View full text Add to dashboard Cite

In biological dosimetry, a radiation dose is estimated using the average number of chromosomal aberration per peripheral blood lymphocytes. To achieve an adequate precision in the estimation of this average, hundreds of cells must be analyzed in 2D microscopy images. Currently, this analysis is performed manually, as conventional computer vision techniques struggle with the wide variety of shapes showcased by chromosomes. The false discovery rate of current automated detection systems is high and variable, depending on small variations in data quality (chromosome spread, illumination variations ...), which makes using it in a fully automated fashion impossible. Automating chromosomal aberration is needed to reduce diagnosis time. Furthermore, an automated system can process more images, which improves confidence intervals around the estimated radiation dose. We build an object detection model to automate chromosomal aberration detection using recent advances in deep convolutional neural networks and statistical learning. We formulated the problem of rare aberration detection as a heatmap regression problem requiring the minimization of a sparsity-promoting loss to reduce the false alarm rate. Our Unet-based approach is analoguous to a one-stage object detector, and keeps the number of hyperparameters to a minimum. Finally, we demonstrate large performance improvements using an ensemble of checkpoints collected during a single run of training.A PCA-based strategy is used to provide cues for interpretation of our deep neural network-based model. The methodology is demonstrated on real, large, and challenging datasets depicting rare chromosomal aberrations and is favorably compared to a reference dosimetry technique.

show abstract

Section: Setting Of Model Parametersmentioning

confidence: 99%

Ensembling Unets, sparse representation and low dimensional visualization for rare chromosomal aberration detection in light microscopy images

Deschemps,

Grégoire,

Martinez

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The experiments in this paper are implemented in a deep learning framework (i.e., PyTorch) on an Nvidia GeForce RTX 2080Ti GPU. We choose the Adam algorithm [66] to optimise our model. The initial learning rate, batch size, and epochs are set as 0.01, 24, and 200.…”

Section: Implementation Detailsmentioning

confidence: 99%

Relationship existence recognition-based social group detection in urban public spaces

Qing

Guo

et al. 2023

Neurocomputing

View full text Add to dashboard Cite

“…In Majeed et al [11] a black-box identificaton (Box-Jenkins) was done for rotary drilling laboratory prototype. Ba et al [12] proposed a methodology for drilling processes monitoring by identifying time varying parameters, where a simple linear system was considered in the identification process.…”

Section: T G Ritto (And)mentioning

confidence: 99%

Bayesian approach to identify the bit–rock interaction parameters of a drill-string dynamical model

Ritto

2014

J Braz. Soc. Mech. Sci. Eng.

View full text Add to dashboard Cite

A drill string is a slender structure used to search for oil and gas. Many works have tackled the problem of modeling the drill-string dynamics in a vertical well. One important aspect in this dynamics is the bit-rock interaction, and, therefore, an identification of the parameters of the bit-rock interaction model becomes crucial. Few works related to this identification problem have been published. The present paper applies the Bayesian approach to identify the parameters of the bit-rock interaction model considering a simplified drill-string dynamical model which takes into account only torsional vibrations. It is assumed an additive Gaussian noise model, and the Metropolis-Hasting algorithm is used to approximate the posterior distribution of the variables analyzed.

show abstract

Stochastic adaptive learning rate in an identification method: An approach for on-line drilling processes monitoring

Cited by 3 publications

References 9 publications

Ensembling Unets, sparse representation and low dimensional visualization for rare chromosomal aberration detection in light microscopy images

Ensembling Unets, sparse representation and low dimensional visualization for rare chromosomal aberration detection in light microscopy images

Relationship existence recognition-based social group detection in urban public spaces

Bayesian approach to identify the bit–rock interaction parameters of a drill-string dynamical model

Contact Info

Product

Resources

About