Rule Preference Effect in Associative Classification Mining

2006 ◽  
Vol 05 (01) ◽  
pp. 13-20 ◽  
Author(s):  
Fadi Thabtah

Classification based on association rule mining, also known as associative classification, is a promising approach in data mining that builds accurate classifiers. In this paper, a rule ranking process within the associative classification approach is investigated. Specifically, two common rule ranking methods in associative classification are compared with reference to their impact on accuracy. We also propose a new rule ranking procedure that adds more tie breaking conditions to the existing methods in order to reduce rule random selection. In particular, our method looks at the class distribution frequency associated with the tied rules and favours those that are associated with the majority class. We compare the impact of the proposed rule ranking method and two other methods presented in associative classification against 14 highly dense classification data sets. Our results indicate the effectiveness of the proposed rule ranking method on the quality of the resulting classifiers for the majority of the benchmark problems, which we consider. This provides evidence that adding more appropriate constraints to break ties between rules positively affects the predictive power of the resulting associative classifiers.

2007 ◽  
Vol 22 (1) ◽  
pp. 37-65 ◽  
Author(s):  
FADI THABTAH

AbstractAssociative classification mining is a promising approach in data mining that utilizes the association rule discovery techniques to construct classification systems, also known as associative classifiers. In the last few years, a number of associative classification algorithms have been proposed, i.e. CPAR, CMAR, MCAR, MMAC and others. These algorithms employ several different rule discovery, rule ranking, rule pruning, rule prediction and rule evaluation methods. This paper focuses on surveying and comparing the state-of-the-art associative classification techniques with regards to the above criteria. Finally, future directions in associative classification, such as incremental learning and mining low-quality data sets, are also highlighted in this paper.


2017 ◽  
Vol 6 (3) ◽  
pp. 71 ◽  
Author(s):  
Claudio Parente ◽  
Massimiliano Pepe

The purpose of this paper is to investigate the impact of weights in pan-sharpening methods applied to satellite images. Indeed, different data sets of weights have been considered and compared in the IHS and Brovey methods. The first dataset contains the same weight for each band while the second takes in account the weighs obtained by spectral radiance response; these two data sets are most common in pan-sharpening application. The third data set is resulting by a new method. It consists to compute the inertial moment of first order of each band taking in account the spectral response. For testing the impact of the weights of the different data sets, WorlView-3 satellite images have been considered. In particular, two different scenes (the first in urban landscape, the latter in rural landscape) have been investigated. The quality of pan-sharpened images has been analysed by three different quality indexes: Root mean square error (RMSE), Relative average spectral error (RASE) and Erreur Relative Global Adimensionnelle de Synthèse (ERGAS).


2005 ◽  
Vol 39 (1) ◽  
pp. 1-26 ◽  
Author(s):  
MARY ARENDS-KUENNING ◽  
FLORA L. KESSY

The low contraceptive prevalence rate and the existence of unmet demand for family planning services present a challenge for parties involved in family planning research in Tanzania. The observed situation has been explained by the demand-side variables such as socioeconomic characteristics and cultural values that maintain the demand for large families. A small, but growing body of research is examining the effect of supply-side factors such as quality of care of family planning services on the demand for contraceptives. This paper analyses the demand and supply factors determining contraceptive use in Tanzania using the Tanzania Service Availability Survey (1996) and the Tanzania Demographic and Health Survey (1996) data sets. The results show that access to family planning services and quality of care of services are important determinants of contraceptive use in Tanzania even after controlling for demand-side factors.


2021 ◽  
Vol 71 (S1) ◽  
pp. 187-203

Abstract This paper examines the factors which determine the impact of network communication and network connections on the likelihood of contracting the new coronavirus in the European and Latin American countries. The author presents several data sets to prove the following suggestions: 1) The generalized indicators of economic development and society’s globalization are not indicators of how vulnerable a country’s population may be in a pandemic; 2) Not the economy as such, but the conventional way of life of people, their daily behaviour and habits have a decisive influence on the disease spread; 3) Factors of prevention of illness and health promotion such as the habit of exercise, distance, and network communications use modern online services to become protective factors against the risk of infection only at a certain level of development of the country; 4) In the developed countries, a much broader set of factors than in the developing countries determine protection against disease risk; 5) The evolution of a networked society opens up significant opportunities for the developing countries to improve the quality of life, and the emergence of new, progressive traditions.


2020 ◽  
pp. 81-93
Author(s):  
D. V. Shalyapin ◽  
D. L. Bakirov ◽  
M. M. Fattakhov ◽  
A. D. Shalyapina ◽  
A. V. Melekhov ◽  
...  

The article is devoted to the quality of well casing at the Pyakyakhinskoye oil and gas condensate field. The issue of improving the quality of well casing is associated with many problems, for example, a large amount of work on finding the relationship between laboratory studies and actual data from the field; the difficulty of finding logically determined relationships between the parameters and the final quality of well casing. The text gives valuable information on a new approach to assessing the impact of various parameters, based on a mathematical apparatus that excludes subjective expert assessments, which in the future will allow applying this method to deposits with different rock and geological conditions. We propose using the principles of mathematical processing of large data sets applying neural networks trained to predict the characteristics of the quality of well casing (continuity of contact of cement with the rock and with the casing). Taking into account the previously identified factors, we developed solutions to improve the tightness of the well casing and the adhesion of cement to the limiting surfaces.


Associative Classification in data mining technique formulates more and more simple methods and processes to find and predict the health problems like diabetes, tumors, heart problems, thyroid, cancer, malaria etc. The methods of classification combined with association rule mining gradually helps to predict large amount of data and also builds the accurate classification models for the future analysis. The data in medical area is sometimes vast and containss the information that relates to different diseases. It becomes difficult to estimate and analyze the disease problems that change from period to period based on severity. In this research paper, the use and need of associative classification for the medical data sets and the application of associative classification on the data in order to predict the by-diseases has been put front. The association rules in this context developed in training phase of data have predicted the chance of occurrence of other diseases in persons suffering with diabetes mellitus using Predictive Apriori. The associative classification algorithms like CAR is deployed in the context of accuracy measures.


2021 ◽  
Vol 126 (4) ◽  
pp. 3321-3336
Author(s):  
T. Liskiewicz ◽  
G. Liskiewicz ◽  
J. Paczesny

AbstractThe citations count is flawed but it still the most common way of measuring the academic impact used by scholarly journals (Impact Factor), individual researchers (h-index) and funding agencies (a proxy for quality of research). Individual papers should attract citations depending upon the importance and usefulness of the results presented. However, large enough data sets reveal that there are parameters independent of individual papers' quality that can determine an average citation rate. Here, we examine papers (4756 in total) published in six selected tribology journals in a six-year window between January 2010 and December 2015. Citations were retrieved from the Web of Science and compared with their (1) manuscript length, (2) number of authors, (3) number of affiliated institutions, (4) number of international co-authors, (5) number of cited references, (6) number of words in the title, and (7) mode of publication (open versus paid access). The results revealed that citations received by papers published in tribology journals are affected by all of these parameters. This is a significant finding for authors wishing to increase the impact of their research. This knowledge can be used effectively at the manuscript planning and writing stages to support scientific merit. We suggest that the significance of parameters not directly related to the quality of a scholarly paper will become more critical with the rise of alternative ways of measuring impact including novel generation of paper metrics (e.g., Eigenfactor, SJR), social mentions, and viral outreach.


2020 ◽  
Vol 27 (2) ◽  
pp. 353-374
Author(s):  
RICARDO P. BEAUSOLEIL

This paper presents an application of Tabu Search algorithm to association rule mining. We focus our attention specifically on classification rule mining, often called associative classification, where the consequent part of each rule is a class label. Our approach is based on seek a rule set handled as an individual. A Tabu search algorithm is used to search for Pareto-optimal rule sets with respect to some evaluation criteria such as accuracy and complexity. We apply a called Apriori algorithm for an association rules mining and then a multiobjective tabu search to a selection rules. We report experimental results where the effect of our multiobjective selection rules is examined for some well-known benchmark data sets from the UCI machine learning repository.


1998 ◽  
Vol 38 (7) ◽  
pp. 777 ◽  
Author(s):  
G. E. Rayment ◽  
K. I. Peverill ◽  
B. C. Shelley

Summary. In relatively few years, the Australian Soil and Plant Analysis Council Inc. (ASPAC) has conducted 2 inter-laboratory proficiency programs on plant material and 3 inter-laboratory proficiency programs on soils. The purpose of these performance-based programs is to enhance the quality of soil and plant analysis in Australasia, with guidance where necessary from the soil and plant expertise of ASPAC members. ASPAC’s inaugural ‘Accreditation Committee’ reviewed published standards and existing laboratory accreditation/proficiency programs in Australia and internationally before developing what is now in full operation. This historical perspective and the 12 principles that guide operations of ASPAC’s soil and plant proficiency programs are described, as are the numeric procedures used to determine satisfactory performance. Certificates are issued to successful laboratories on completion of each program. Moreover, these remain current until signed certificates from the next equivalent program are released. Wide variations in some data sets suggest there is considerable scope to improve laboratory accuracy, particularly for soil chemical tests. Some of these differences are sufficient to markedly affect the assessment of fertiliser requirements. The present ‘Accreditation Committee’, in addition to State Representatives, serve as ‘points-of-contact’ for laboratories that require assistance to overcome problems with analytical accuracy and precision. ASPAC encourages its member laboratories to seek and maintain NATA (National Association of Testing Authorities, Australia) accreditation, in addition to participating regularly in the performance-based proficiency programs run by ASPAC.


Sign in / Sign up

Export Citation Format

Share Document