scholarly journals PERFORMANCE DA MODELAGEM PARA CLASSIFICAÇÃO DE SÍTIOS FLORESTAIS EM BASES DE DADOS COM OUTLIERS

Nativa ◽  
2021 ◽  
Vol 9 (1) ◽  
pp. 54-61
Author(s):  
Pábulo Diogo de Souza ◽  
Carlos Alberto Araújo Júnior ◽  
Christian Dias Cabacinha ◽  
Leandro Silva de Oliveira ◽  
Celso Dotta Lopes Junior ◽  
...  
Keyword(s):  
Radial Basis Function ◽  
Basis Function ◽  
Multilayer Perceptron ◽  
Forest Inventory ◽  
Eucalyptus Grandis ◽  
Site Index ◽  
Site Classification ◽  
Forest Site ◽  
Forest Sites ◽  
Forest Site Classification

As informações utilizadas para estimativa da capacidade produtiva de sítios florestais provêm de bases de dados de inventário florestal que podem conter observações discrepantes (outliers). Assim, torna-se necessário a análise de consistência para exclusão destes. Porém, os outliers podem representar determinado padrão de crescimento existente na floresta, logo a exclusão destes pode ser uma ação equivocada. Objetivou-se comparar a performance de diferentes técnicas de modelagem para classificação de sítios florestais, considerando uma base de dados com a presença de outliers. Utilizou-se pares de dados de idade e altura dominante (HD) de parcelas permanentes de Eucalyptus urophyla x Eucalyptus grandis localizadas no norte de Minas Gerais. Foi simulado um outlier de HD. A base de dados foi modelada, com e sem presença de outliers, por regressão linear (RL) e redes neurais artificiais Multilayer Perceptron (MLP) e Radial Basis Function (RBF). Os métodos foram analisados por meio dos critérios estatísticos de precisão: bias, raiz quadrada do erro médio, correlação de Pearson, erro médio percentual e gráfico de dispersão residual. A MLP foi superior para estimativa do índice de sítio. Portanto, a MLP é indicada para classificação de sítios florestais quando há presença de outliers na base de dados. Palavras-chave: índice de sítio; inventário florestal; dados discrepantes.   Performance of modeling for classification of forest sites in databases with outliers   ABSTRACT: The information used to estimate the productive capacity of forest sites comes from forest inventory databases that may contain discrepant observations (outliers). Thus, consistency analysis is required to exclude these. However, the outliers may represent a certain growth pattern existing in the forest, so their exclusion may be a mistaken action. The objective was to compare the performance of different modeling techniques for forest site classification, considering a database with the presence of outliers. We used pairs of data of age and dominant height (HD) of permanent parcels of Eucalyptus urophila x Eucalyptus grandis located in the north of Minas Gerais. A HD outlier was simulated. The database was modeled, with and without the presence of outliers, by linear regression (RL) and artificial neural networks Multilayer Perceptron (MLP) and Radial Basis Function (RBF). The methods were analyzed by means of precision statistical criteria: bias, square root of mean error, Pearson correlation, mean percentage error and residual scatter plot. The MLP was superior for site index estimation. Therefore, the MLP is indicated for forest site classification when there are outliers in the database. Keywords: site index; forest inventory; discrepant data.

An ecological land classification approach to modeling the production of forest biomass

2011 ◽  
Vol 87 (1) ◽  
pp. 23-32 ◽  
Author(s):  
Bharat Pokharel ◽  
Jeffery P Dech
Keyword(s):  
Forest Management ◽  
Forest Biomass ◽  
Site Index ◽  
Site Classification ◽  
Forest Site ◽  
Land Classification ◽  
Site Productivity ◽  
Ecological Land Classification ◽  
Site Evaluation ◽  
Forest Site Classification

Forest site classification is a prerequisite to successful integrated forest resources planning and management. Traditionally,site classification has emphasized a phytocentric approach, with tools such as the site index having a rich and longhistory in forest site evaluation. The concept of site index was primarily devised to assess site productivity of an even-aged,single-species stand. Site index has been the primary method of forest site evaluation in support of management for traditionalforest products. However, this method of site classification has been criticized as the needs, perspectives andsocial values of the public regarding forest management have changed the emphasis from timber production to multiplevalueforestry practices. There are alternative approaches to forest site classification that have the potential to meet thegrowing demands placed on forest information for inventory and modeling purposes. Ecological Land Classification(ELC), is a phytogeocentric approach that stratifies the landscape into ecologically meaningful units (ecosites) based onsubstrate characteristics, moisture regime and canopy composition. This approach offers a more holistic view of site productivityevaluation; however, until recently it has been difficult to acquire data to support widespread mapping ofecosites. Remote sensing technology along with predictive modeling and interpretive mapping techniques make the applicationof an ecosite-based approach at the forest landscape level possible. As forest management moves towards the considerationof a broader set of resources (e.g., woody biomass), there is an opportunity to develop new tools for linking forestproductivity to the sustainable production of forest bioproducts with forest ecosites as a solid foundation forsegmenting the landscape. Key words: forest site classification, site index, site productivity, Ecological Land Classification (ELC), ecosites, forest biomass,bioproducts

The current status of forest site classification in Ontario

1992 ◽  
Vol 68 (1) ◽  
pp. 64-77 ◽  
Author(s):  
R. A. Sims ◽  
P. Uhlig
Keyword(s):  
Resource Management ◽  
Building Blocks ◽  
Current Status ◽  
Background Information ◽  
Site Classification ◽  
Land Resource ◽  
Forest Site ◽  
Forest Sites ◽  
Soil Site ◽  
Forest Site Classification

Forest sites are diagnostic forest-landscape ecosystem units that resource managers must deal with during the planning and implementation stages of management. Forest sites are the basic building blocks for undertaking integrated resource management which weighs wildlife, recreation, environmental impact and various other concerns along with timber harvesting. Consequently, accurate and practical systems for classifying and mapping forest sites are becoming increasingly necessary to organize, communicate and use existing and new management knowledge and experience effectively.Over the past four decades in Ontario, a number of studies and resource surveys have provided important background information on forest sites. Many have considered, to varying extents, the integrative roles of vegetation, soil-site, landform and general climate on forests and forest land. Generally, the emphasis has been on description and classification, with results generating a better understanding of how various forests in different areas develop, both qualitatively and quantitatively, in relation to soil-site or other features of the basic land resource. Some of these studies and surveys have been instrumental in advancing the definitions and understanding of forested ecosystems. Others have provided new information on site dynamics, interrelationships and functions, or have contributed to the science (and art) of site evaluation and classification.This paper briefly summarizes the current status of forest site classification in Ontario. Over time, the role of forest site classification has evolved in response to new technologies and information, and to new emphases and values in resource management. In general, site classification research has become increasingly integrative and quantitative. Some of the important future challenges facing forest site classification in Ontario are briefly discussed. Key words: ecological land classification, forest ecology, forest ecosystem classification, forest management interpretations, forest site classification, land use planning, Ontario.

scholarly journals Estudo sobre Métricas para Definir Reputação do Autor de Comentários em Sites de Vendas de Produtos

2018 ◽  
Author(s):  
Carlos Augusto De Sá ◽  
Raimundo Santos Moura
Keyword(s):  
Radial Basis Function ◽  
Basis Function ◽  
Multilayer Perceptron ◽  
Radial Basis

Conhecer a reputação do autor de textos opinativos é de suma importância para avaliação de comentários na Web. Este artigo apresenta um estudo sobre medidas usadas no processo de avaliação da reputação do autor em sites de vendas de produtos. Realizou-se dois experimentos com as redes neurais Multilayer Perceptron (MLP) e Radial Basis Function (RBF), sendo que a rede MLP obteve melhor desempenho. Comparou-se também a abordagem TOP(X) original, usada para inferir os melhores comentários, com um novo modelo que utiliza rede MLP na dimensão da reputação do autor. Considerando os comentários excelentes e bons, a nova abordagem apresentou resultados significativamente superiores.

scholarly journals Québec forest ecological classification program

1992 ◽  
Vol 68 (1) ◽  
pp. 53-63 ◽  
Author(s):  
Jean-François Bergeron ◽  
Jean-Pierre Saucier ◽  
Denis Robert ◽  
André Robitaille
Keyword(s):  
Multivariate Analysis ◽  
Physical Environment ◽  
Forest Inventory ◽  
Forest Ecology ◽  
Site Classification ◽  
Forest Site ◽  
Ecological Classification ◽  
Forest Types ◽  
Ecological Data ◽  
Surface Deposit

In 1986, the ministère des Forêts du Québec instituted a provincial program to study forest ecosystems entitled the "Forest Ecological Classification (FEC) Program." Under this program, a multidisciplinary team was charged with conducting ecological surveys, analyzing and characterizing the variables of the physical environment, classifying vegetation and preparing integrated forest inventory maps. Their goal is to complete the ecological classification of the forests in all territories south of the 52nd parallel. To undertake such a vast project, it was necessary to prepare detailed methodological guides for data collection, data analysis and mapping. The following products are now available for many different ecological regions: classifications of forest types, toposequences, physiographic and surface deposit maps and integrated forest inventory maps. Multivariate analysis methods are used in analyzing ecological data; in this way, hierarchical classifications and ordinations can be used as the basis for identifying and describing forest types, vegetation-physical environment relationships and successional patterns. Such ecological classification products are an indispensable tool for forest managers and users. Key words: ecological classification, forest ecology, forest management, forest site classification, multivariate analysis, physical environment, Québec.

Fast Learning in Neural Networks

2008 ◽  
pp. 91-104
Author(s):  
Darryl Charles ◽  
Colin Fyfe ◽  
Daniel Livingstone ◽  
Stephen McGlinchey
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Continuous Function ◽  
Radial Basis Function ◽  
Supervised Learning ◽  
Basis Function ◽  
Multilayer Perceptron ◽  
Rbf Networks ◽  
Rbf Network ◽  
Fast Learning

We noted in the previous chapters that, while the multilayer perceptron is capable of approximating any continuous function, it can suffer from excessively long training times. In this chapter we will investigate methods of shortening training times for artificial neural networks using supervised learning. (Haykin, 1999) is a particularly good reference for radial basis function, RBF, networks. In this chapter we outline the theory and implementation of a RBF network before demonstrating how such a network may be used to solve one of the previously visited problems, and compare our solutions.

Critical heat flux prediction by using radial basis function and multilayer perceptron neural networks: A comparison study

2007 ◽  
Vol 237 (4) ◽  
pp. 377-385 ◽  
Author(s):  
Nima Vaziri ◽  
Alireza Hojabri ◽  
Ali Erfani ◽  
Mehrdad Monsefi ◽  
Behnam Nilforooshan
Keyword(s):  
Neural Networks ◽  
Heat Flux ◽  
Radial Basis Function ◽  
Critical Heat Flux ◽  
Basis Function ◽  
Multilayer Perceptron ◽  
Comparison Study ◽  
Radial Basis
Sign in / Sign up

Export Citation Format

Share Document