scholarly journals Experimental method to predict avalanches based on neural networks

2016 ◽  
Vol 56 (4) ◽  
pp. 502-510 ◽  
Author(s):  
V. V. Zhdanov
Keyword(s):  
Neural Networks ◽  
Training Sample ◽  
Model Quality ◽  
Linear Discriminant ◽  
Canonical Variables ◽  
Russian Version ◽  
The Neural Networks ◽  
Trained Neural Network ◽  
Experimental Use

The article presents results of experimental use of currently available statistical methods to classify the avalanche‑dangerous precipitations and snowfalls in the Kishi Almaty river basin. The avalanche service of Kazakhstan uses graphical methods for prediction of avalanches developed by I.V. Kondrashov and E.I. Kolesnikov. The main objective of this work was to develop a modern model that could be used directly at the avalanche stations. Classification of winter precipitations into dangerous snowfalls and non‑dangerous ones was performed by two following ways: the linear discriminant function (canonical analysis) and artificial neural networks. Observational data on weather and avalanches in the gorge Kishi Almaty in the gorge Kishi Almaty were used as a training sample. Coefficients for the canonical variables were calculated by the software «Statistica» (Russian version 6.0), and then the necessary formula had been constructed. The accuracy of the above classification was 96%. Simulator by the authors L.N. Yasnitsky and F.М. Cherepanov was used to learn the neural networks. The trained neural network demonstrated 98% accuracy of the classification. Prepared statistical models are recommended to be tested at the snow‑avalanche stations. Results of the tests will be used for estimation of the model quality and its readiness for the operational work. In future, we plan to apply these models for classification of the avalanche danger by the five‑point international scale.

Automate Labeling Of Bugs and Tickets Using Attention-Based Mechanism in Recurrent Neral Networks

2021 ◽  
Vol 9 (11) ◽  
pp. 1411-1418
Author(s):  
Ravi Kauthale
Keyword(s):  
Neural Networks ◽  
Text Classification ◽  
Recurrent Neural Networks ◽  
Support Systems ◽  
The Neural Networks

Abstract: The aim here is to explore the methods to automate the labelling of the information that is present in bug trackers and client support systems. This is majorly based on the classification of the content depending on some criteria e.g., priority or product area. Labelling of the tickets is important as it helps in effective and efficient handling of the ticket and help is quicker and comprehensive resolution of the tickets. The main goal of the project is to analyze the existing methodologies used for automated labelling and then use a newer approach and compare the results. The existing methodologies are the ones which are based of the neural networks and without neural networks. In this project, a newer approach based on the recurrent neural networks which are based on the hierarchical attention paradigm will be used. Keywords: Automate Labeling, Recurrent Neural Networks, Hierarchical Attention, Multi-class Text Classification, GRU

scholarly journals Toward automated classification of monolayer versus few-layer nanomaterials using texture analysis and neural networks

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Shrouq H. Aleithan ◽  
Doaa Mahmoud-Ghoneim
Keyword(s):  
Neural Networks ◽  
Texture Analysis ◽  
Higher Order ◽  
Image Texture ◽  
Proper Class ◽  
Automated Classification ◽  
Linear Discriminant ◽  
Ann Classifier ◽  
Occurrence Matrix

AbstractThe need for a fast and robust method to characterize nanostructure thickness is growing due to the tremendous number of experiments and their associated applications. By automatically analyzing the microscopic image texture of MoS2 and WS2, it was possible to distinguish monolayer from few-layer nanostructures with high accuracy for both materials. Three methods of texture analysis (TA) were used: grey level histogram (GLH), grey levels co-occurrence matrix (GLCOM), and run-length matrix (RLM), which correspond to first, second, and higher-order statistical methods, respectively. The best discriminating features were automatically selected using the Fisher coefficient, for each method, and used as a base for classification. Two classifiers were used: artificial neural networks (ANN), and linear discriminant analysis (LDA). RLM with ANN was found to give high classification accuracy, which was 89% and 95% for MoS2 and WS2, respectively. The result of this work suggests that RLM, as a higher-order TA method, associated with an ANN classifier has a better ability to quantify and characterize the microscopic structure of nanolayers, and, therefore, categorize thickness to the proper class.

Neural networks application for phasechronometric measurement information processing

2020 ◽  
pp. 31-35
Author(s):  
D. D. Boldasov ◽  
J. V. Drozdova ◽  
A. S. Komshin ◽  
A. B. Syritskii
Keyword(s):  
Neural Networks ◽  
Information Processing ◽  
Binary Classification ◽  
Processing Technique ◽  
Measurement Information ◽  
Perceptron Algorithm ◽  
Classification Feature ◽  
Operation Modes ◽  
The Neural Networks

This article describes the processing technique of measuring phasechronometric information based on the neural networks use. The novelty of the proposed approach lies in the choice of a classification feature and the perceptron algorithm use as an algorithm for binary classification performing. In this article, to assess the concept operability, the simplest binary classification of the lathe operation modes is made: idle or cutting.

Lettuce Images Features Extraction and Intelligent Classification of Growth Period

2013 ◽  
Vol 846-847 ◽  
pp. 1351-1354
Author(s):  
Yan Wang ◽  
Jun Sun ◽  
Hong Li Wang ◽  
Ze Gao Dai ◽  
Wen Xia Lv
Keyword(s):  
Neural Networks ◽  
Cluster Analysis ◽  
Recognition Accuracy ◽  
Growth Period ◽  
Modeling And Analysis ◽  
Good Image ◽  
Visible Images ◽  
The Neural Networks ◽  
Lettuce Growth

Classification of lettuce growth peroid is the premise of records of lettuce growth information. In this study, lettuce images in every growth period are collected. And visible images are preprocessed to extract features to establish initial feature library of lettuce images. Through R cluster analysis on many features, good image eigenvector are obtained. Classification of the lettuce samples are obtained by modeling and analysis of the neural networks. The experimental classification results compared with practical classification results, the recognition accuracy is up to 88.4%.

scholarly journals CONCEPTUAL BASES OF MACRO PREDICTION ON THE BASIS OF THE NEURAL NETWORKS SYSTEMS

2017 ◽  
pp. 68-73
Author(s):  
Inna Strelchenko
Keyword(s):  
Neural Networks ◽  
Information Exchange ◽  
Large Scale ◽  
System Analysis ◽  
Global Financial Crisis ◽  
Functional Model ◽  
Training Sample ◽  
Sharp Change ◽  
The Neural Networks ◽  
The Global Financial Crisis

Introduction. Under the conditions of accelerated integration of global financial markets, the growth of information exchange speed and large-scale transmission of electronic money, the forecasting of economy scenarios under a sharp change in the environment, particularly during the global financial crisis becomes an important task. Purpose. The aim of the research is to develop and substantiate the conceptual principles of macro-prediction on the basis of the neural networks system in conditions of increasing non-linearity of the environment. Method (methodology). The use of system analysis, integrated approaches which are based on the principles of economic theory, on the one hand, and, on the other, on the conceptual basis of economic and mathematical modeling of economics and finances processes, have become the methodological basis of the research. Results. The use of the neural networks system in the task of macro prediction of the economic systems behavior under the crisis conditions has been justified. It has been definitedthe list of indicators that formed a training sample for simulation. The basic functional model of information technology of the forecasting process has been constructed.

scholarly journals 10.5937/industrija42-5686 = The implementation of the neural networks to the problem of economic classification of countries

Industrija ◽  
2014 ◽  
Vol 42 (4) ◽  
pp. 25-42
Author(s):  
Sasa Obradovic ◽  
Miljan Lekovic ◽  
Milos Marinkovic
Keyword(s):  
Neural Networks ◽  
The Neural Networks

scholarly journals The classification of solvents based on solvatochromic characteristics: the choice of optimal parameters for artificial neural networks

2012 ◽  
Vol 10 (4) ◽  
pp. 1318-1327
Author(s):  
Yaroslava Pushkarova ◽  
Yuriy Kholin
Keyword(s):  
Neural Networks ◽  
Hydrogen Bond ◽  
Artificial Neural Networks ◽  
Supervised Classification ◽  
Solvatochromic Parameters ◽  
Optimal Parameters ◽  
Linear Discriminant ◽  
Solvatochromic Probes ◽  
Artificial Neural

AbstractThe Taft-Kamlet-Abboud hydrogen-bond acidity, hydrogen-bond basicity and polarity-polarizability are widely used as empirical characteristics of solvent-solute interactions. These solvatochromic parameters are determined from the absorption band positions of solvatochromic probes in the standard medium and in the medium under study. The practice of solvatochromic probing is growing rapidly, and the values of solvatochromic parameters are refined from time to time. As these values are rather close for many media, the classification of media based on these values can be tedious. This increases the choice of algorithms that can be employed in order to decrease the ambiguity of classification. The classification algorithms stable to small variations of solvatochromic parameters are of special interest. The artificial neural networks (ANN) proved to be a powerful tool for the supervised classification. The paper focuses on the search of optimal parameters of probabilistic, dynamic, Elman, feed-forward, and cascade ANN for the classification of solvent on the basis of their solvatochromic characteristics. Also, the influence of data variation on the stability of classification is examined. The dynamic and probabilistic neural networks have been found to be error-free and stable; they have significantly become such a common tool for supervised classification as linear discriminant analysis.

A Fault Diagnosis Method Based on Combination of Neural Network and Fault Dictionary

2013 ◽  
Vol 765-767 ◽  
pp. 2078-2081
Author(s):  
Ya Feng Meng ◽  
Sai Zhu ◽  
Rong Li Han
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Fault Diagnosis ◽  
Real Time ◽  
Large Scale ◽  
Fault Dictionary ◽  
Diagnosis Method ◽  
The Neural Networks ◽  
Trained Neural Network ◽  
Complex Circuits

Neural network and Fault dictionary are two kinds of very useful fault diagnosis method. But for large scale and complex circuits, the fault dictionary is huge, and the speed of fault searching affects the efficiency of real-time diagnosing. When the fault samples are few, it is difficulty to train the neural network, and the trained neural network can not diagnose the entire faults. In this paper, a new fault diagnosis method based on combination of neural network and fault dictionary is introduced. The fault dictionary with large scale is divided into several son fault dictionary with smaller scale, and the search index of the son dictionary is organized with the neural networks trained with the son fault dictionary. The complexity of training neural network is reduced, and this method using the neural networks ability that could accurately describe the relation between input data and corresponding goal organizes the index in a multilayer binary tree with many neural networks. Through this index, the seeking scope is reduced greatly, the searching speed is raised, and the efficiency of real-time diagnosing is improved. At last, the validity of the method is proved by the experimental results.

Sign in / Sign up

Export Citation Format

Share Document