Please use this identifier to cite or link to this item:
|Title:||An Assessment of Classification with Hybrid Methodology for Neural Network Classifier against different Classifiers|
Singh Rathore, Pramod
|Keywords:||Naive bayes classifier; Simple cart; Rule classification; Trees Classifier|
|Description:||Introduction: The research An Assessment of Classification with Hybrid Methodology for Neural Network Classifier against different Classifiers focuses on working efficiency of different classifiers and comparative results. Authentic & dynamic datasets are considered in order to get at specific meaning full results and was conducted at M.D.S University Ajmer in 2017.Methods: According to the Indian Heart Association, almost half of heart attack victim Indian citizen in enduring this disease by age of 50 years and around one-fourth of all heart-attack happen in Indian occur under 40 years of age. In the considerable amount of cases, lack of cognition is the major reason for the delay in treatment.Results: So we are giving a prototype definition of a Hybrid Technology of Analyzing Accuracy measure of neural network classifiers with Bayesian classifiers, Rules-based classifiers and tree-based classifiers on Diagnosis of Heart Disease dataset.Conclusions: This Methodology will distinguish the functioning, as well as the execution results of these data mining classifiersOriginality: The work consider large dataset and different approach has been applied classifier function naïve Bayes algorithm, is taken as base algorithm and all other algorithms like, Rule classification algorithm Ridor and Trees Classifier algorithm Simple Cart are used for comparison purpose, in order to predict angiographic disease status of patient. Limitations: As the proposed work is new implementation it is needed to be implemented on different dynamic datasets to get assured for result accuracy and efficiency of proposed method.|
|Publisher:||Universidad Cooperativa de Colombia|
|Appears in Collections:||Revista Ingeniería Solidaria|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.