The application of optimal weights initialization algorithm based on K-L transform in multi-layer perceptron networks

2013 ◽  
Author(s):  
Wei Xiao ◽  
Dun Pu ◽  
Zhicheng Dong ◽  
Cungen Liu
Keyword(s):  
Multi Layer Perceptron ◽  
Optimal Weights ◽  
Initialization Algorithm

Q-optimal experimental designs and close to them experimental designs for polynomial regression on the interval

2020 ◽  
Vol 86 (5) ◽  
pp. 65-72
Author(s):  
Yu. D. Grigoriev
Keyword(s):  
Optimal Design ◽  
Polynomial Regression ◽  
Direct Consequence ◽  
Experimental Designs ◽  
Optimal Designs ◽  
Software Systems ◽  
General Character ◽  
Support Points ◽  
Optimal Weights ◽  
Universal Expression

The problem of constructing Q-optimal experimental designs for polynomial regression on the interval [–1, 1] is considered. It is shown that well-known Malyutov – Fedorov designs using D-optimal designs (so-called Legendre spectrum) are other than Q-optimal designs. This statement is a direct consequence of Shabados remark which disproved the Erdős hypothesis that the spectrum (support points) of saturated D-optimal designs for polynomial regression on a segment appeared to be support points of saturated Q-optimal designs. We present a saturated exact Q-optimal design for polynomial regression with s = 3 which proves the Shabados notion and then extend this statement to approximate designs. It is shown that when s = 3, 4 the Malyutov – Fedorov theorem on approximate Q-optimal design is also incorrect, though it still stands for s = 1, 2. The Malyutov – Fedorov designs with Legendre spectrum are considered from the standpoint of their proximity to Q-optimal designs. Case studies revealed that they are close enough for small degrees s of polynomial regression. A universal expression for Q-optimal distribution of the weights pi for support points xi for an arbitrary spectrum is derived. The expression is used to tabulate the distribution of weights for Malyutov – Fedorov designs at s = 3, ..., 6. The general character of the obtained expression is noted for Q-optimal weights with A-optimal weight distribution (Pukelsheim distribution) for the same problem statement. In conclusion a brief recommendation on the numerical construction of Q-optimal designs is given. It is noted that in this case in addition to conventional numerical methods some software systems of symbolic computations using methods of resultants and elimination theory can be successfully applied. The examples of Q-optimal designs considered in the paper are constructed using precisely these methods.

Aplicação de Multi-Layer Perceptron para Previsão de Emissão de Gases derivados de Veículos a Diesel

2017 ◽  
Vol 3 (2) ◽  
pp. 1-11
Author(s):  
Luís Otávio Rigo Jr. ◽  
Jesuina Cássia Santiago de Araújo ◽  
Leandro Nogueira dos Santos ◽  
Mona Lisa Moura de Oliveira
Keyword(s):  
Selective Catalytic Reduction ◽  
Catalytic Reduction ◽  
Back Propagation ◽  
Multi Layer Perceptron

Fontes veiculares movidos a Diesel têm contribuido significativamente para o aumento da poluição atmosférica. A tendência mundial de utilizar motor Diesel se deve ao rendimento real alcançado por esta máquina (~34%), quando comparada com motores Otto (~26%). Em termos de poluição, tais motores apresentam a vantagem de emitir menor concentração de hidrocarbonetos e CO2. Por outro lado, o motor Diesel apresenta a desvantagem de emitir materiais particulados e NOx. Com fins de atender a legislação, tem sido incorporado aos veículos a Diesel um sistema catalítico, que injeta uréia nos gases de escape. Tal processo, conhecido como SCR (Selective Catalytic Reduction), tem por finalidade transformar NOx em N2 e H2O. Órgãos governamentais têm atuado como agentes controladores, exigindo dos fabricantes de motores soluções tecnológicas, capazes de reduzir os níveis de emissões destes poluentes. Essas soluções estão atreladas a uma série de testes experimentais onerosos. Tendo-se em vista que as taxas de emissão de NOx dependem de fatores que se correlacionam de forma complexa, faz-se necessário à utilização de ferramentas de simulação para prever tais taxas. Neste trabalho, foi utilizada uma Rede Neural Artificial, denominada Multi-Layer Perceptron, com algoritmo de aprendizado supervisionado Back Propagation, para estimar as taxas de emissão dos gases NOx, NH3 e N2O em veículos a Diesel. Os resultados mostraram que parâmetros de entrada (velocidade espacial, temperatura, concentração de NOx, de NH3, de O2 , de SO2 e de H2O) se correcionam fortemente com as taxas de emissão de NOx e NH3 na saída. Este fator foi comprovado pela grande capacidade de aprendizado das redes testadas, com erro médio próximo de 0,01 no conjunto de aprendizado. Os resultados sobre o conjunto de teste demonstraram, também, grande capacidade de generalização das redes. O melhor resultado encontrado foi de 2,9% para NOx e NH3 e de N2O de 5,1%. Estes resultados revelam que a RNA demonstrou ser um método eficiente para prever as taxas de emissão de poluentes em perímetro urbano e rodovias.

A neural computing-based cash price prediction using multi-layer perceptron (MLP) and feature selection for health economics

2019 ◽  
Vol 14 ◽  
Author(s):  
Zinat Ansari
Keyword(s):  
Feature Selection ◽  
Health Economics ◽  
Cash Flow ◽  
Multi Layer Perceptron ◽  
Asset Sales ◽  
Fixed Asset ◽  
Academic Fields ◽  
Investment Returns ◽  
Short And Long Term

Background: The present study proceeds to incorporate feature selection as a means for selecting the most relevant features affecting the prediction of cash prices in Iran in terms of health economics. Health economics are between academic fields that can aid in ameliorating conditions so as to perform better decisions in regards to the economy such as determining cash prices. Methods: Accordingly, a series of search algorithms, namely the Best-First, Greedy-Stepwise, and Ranker methods, are deployed in order to extract the most relevant features from among a 500 data samples. The validity of the methods was evaluated via the LMT procedure. The corresponding dataset used for this study constitutes a variety of features including net cash flow, dividends, revenue from short and long-term deposits, cash flow from investment returns, income tax, fixed asset purchases, fixed asset sales, long-term investment purchases, long-term investment sales, total cash flow from investment activities, financial facilities, and repayment of financial facilities. Results: The results were indicative of the superiority of the Ranker model using the RelieF-Attribute-Eval tool in Weka over the remaining classification methods. Ergo, the LMT approach could be employed to remove data redundancies and thereby accelerate the estimation process, while saving time and money. The results of the multi-layer perceptron (MLP) further confirmed the high accuracy of the proposed method in estimating cash prices. Conclusions: The present research attempted to reduce the volume of data required for predicting end cash by means of employing a feature selection so as to save both precious money and time.

scholarly journals The Classification of Medicinal Plant Leaves Based on Multispectral and Texture Feature Using Machine Learning Approach

Agronomy ◽  
2021 ◽  
Vol 11 (2) ◽  
pp. 263
Author(s):  
Samreen Naeem ◽  
Aqib Ali ◽  
Christophe Chesneau ◽  
Muhammad H. Tahir ◽  
Farrukh Jamal ◽  
...  
Keyword(s):  
Machine Learning ◽  
Medicinal Plant ◽  
Texture Feature ◽  
Stevia Rebaudiana ◽  
Ocimum Sanctum ◽  
Multi Layer Perceptron ◽  
Plant Leaves ◽  
Chi Square ◽  
Lemon Balm

This study proposes the machine learning based classification of medical plant leaves. The total six varieties of medicinal plant leaves-based dataset are collected from the Department of Agriculture, The Islamia University of Bahawalpur, Pakistan. These plants are commonly named in English as (herbal) Tulsi, Peppermint, Bael, Lemon balm, Catnip, and Stevia and scientifically named in Latin as Ocimum sanctum, Mentha balsamea, Aegle marmelos, Melissa officinalis, Nepeta cataria, and Stevia rebaudiana, respectively. The multispectral and digital image dataset are collected via a computer vision laboratory setup. For the preprocessing step, we crop the region of the leaf and transform it into a gray level format. Secondly, we perform a seed intensity-based edge/line detection utilizing Sobel filter and draw five regions of observations. A total of 65 fused features dataset is extracted, being a combination of texture, run-length matrix, and multi-spectral features. For the feature optimization process, we employ a chi-square feature selection approach and select 14 optimized features. Finally, five machine learning classifiers named as a multi-layer perceptron, logit-boost, bagging, random forest, and simple logistic are deployed on an optimized medicinal plant leaves dataset, and it is observed that the multi-layer perceptron classifier shows a relatively promising accuracy of 99.01% as compared to the competition. The distinct classification accuracy by the multi-layer perceptron classifier on six medicinal plant leaves are 99.10% for Tulsi, 99.80% for Peppermint, 98.40% for Bael, 99.90% for Lemon balm, 98.40% for Catnip, and 99.20% for Stevia.

Multi-layer Perceptron based Photovoltaic Forecasting for Rooftop PV Applications in Smart Grid

2020 ◽  
Author(s):  
Imtiaz Parvez ◽  
Arif Sarwat ◽  
Anjan Debnath ◽  
Temitayo Olowu ◽  
Md Golam Dastgir ◽  
...  
Keyword(s):  
Smart Grid ◽  
Multi Layer Perceptron
Sign in / Sign up

Export Citation Format

Share Document