Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: batch normalization

Sortuj według:

Ogranicz wyniki do:

A robust inversion of logging-while-drilling responses based on deep neural network

Zhu Gaoyang, Gao Muzhi, Wang Bin

Acta Geophysica

2024

Vol. 72, no. 1

129--139

Resistivity inversion plays a significant role in recent geological exploration, which can obtain formation information through logging data. However, resistivity inversion faces various challenges in practice. Conventional inversion approaches are always time-consuming, nonlinear, non-uniqueness, and ill-posed, which can result in an inaccurate and inefficient description of subsurface structure in terms of resistivity estimation and boundary location. In this paper, a robust inversion approach is proposed to improve the efficiency of resistivity inversion. Specifically, inspired by deep neural networks (DNN) remarkable nonlinear mapping ability, the proposed inversion scheme adopts DNN architecture. Besides, the batch normalization algorithm is utilized to solve the problem of gradient disappearing in the training process, as well as the k-fold cross-validation approach is utilized to suppress overfitting. Several groups of experiments are considered to demonstrate the feasibility and efficiency of the proposed inversion scheme. In addition, the robustness of the DNN-based inversion scheme is validated by adding different levels of noise to the synthetic measurements. Experimental results show that the proposed scheme can achieve faster convergence and higher resolution than the conventional inversion approach in the same scenario. It is very significant for geological exploration in layered formations.

Selected technical issues of deep neural networks for image classification purposes

Grochowski Michał, Kwasigroch A., Mikołajczyk A.

Bulletin of the Polish Academy of Sciences. Technical Sciences

2019

Vol. 67, nr 2

363--376

In recent years, deep learning and especially deep neural networks (DNN) have obtained amazing performance on a variety of problems, in particular in classification or pattern recognition. Among many kinds of DNNs, the convolutional neural networks (CNN) are most commonly used. However, due to their complexity, there are many problems related but not limited to optimizing network parameters, avoiding overfitting and ensuring good generalization abilities. Therefore, a number of methods have been proposed by the researchers to deal with these problems. In this paper, we present the results of applying different, recently developed methods to improve deep neural network training and operating. We decided to focus on the most popular CNN structures, namely on VGG based neural networks: VGG16, VGG11 and proposed by us VGG8. The tests were conducted on a real and very important problem of skin cancer detection. A publicly available dataset of skin lesions was used as a benchmark. We analyzed the influence of applying: dropout, batch normalization, model ensembling, and transfer learning. Moreover, the influence of the type of activation function was checked. In order to increase the objectivity of the results, each of the tested models was trained 6 times and their results were averaged. In addition, in order to mitigate the impact of the selection of learning, test and validation sets, k-fold validation was applied.