A survey on geographic classification of virgin olive oil with using T-operators in Fuzzy Decision Tree Approach


Nasibov E., Savas S. K., VAHAPLAR A., KINAY A. Ö.

CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, vol.155, pp.86-96, 2016 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 155
  • Publication Date: 2016
  • Doi Number: 10.1016/j.chemolab.2016.04.004
  • Journal Name: CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.86-96
  • Keywords: Pattern classification, Machine learning, Fuzzy ID3 algorithm, T-operators, Geographic classification, NEURAL-NETWORKS, GENERAL-CLASS, NORMS, SETS, CONNECTIVES, PERFORMANCE, INDUCTION
  • Dokuz Eylül University Affiliated: Yes

Abstract

Olive oil is a crucial agricultural food product from past to present. The quality control of this product is too difficult. The geographic classification of it has an importance for the countries in order to provide the traceability. This paper aims to present a classification system for the geographic classification of virgin olive oil based on chemical parameters which contain uncertainty. Proposed system constructs the rules by using fuzzy decision tree algorithm. This algorithm builds rules by using ID3 algorithm with fuzzy entropy on the fuzzified data. The reasoning procedure based on rule based classification is handled with different T-operators. Fuzzy c-means algorithm is used in order to fuzzify the olive oil data set. The cluster numbers of each variable are decided according to partition coefficient validity criteria. The model is examined by using different decision tree approaches (C4.5 and standard version Fuzzy ID3 algorithm). The quality of proposed FID3 reasoning method with nine different T-operators is analyzed by using accuracy rates handled with 20 threshold values. Also, the conclusions are supported by statistical analysis. Experimental results show that fuzzy reasoning method has a crucial manner for the geographic classification. The classification system can perform better performance via different parameters for parametric T-operators. (C) 2016 Elsevier B.V. All rights reserved.