Abstract
In this investigation, an approach using Coac-tive Neuro-Fuzzy Inference System (CANFIS) as diagnosis system for breast cancer has been proposed on Wisconsin Breast Cancer Data (WBCD). It is occasionally difficult to attain the ultimate diagnosis even for medical experts due to the complexity and non-linearity of the rela-tionships between the large measured factors, which can be possibly resolved with a human like decision-making process using Artificial Intelligence (AI) algorithms. CANFIS is an AI algorithm which has the advantages of both fuzzy inference system and neural networks and can deal with ambiguous data and learn from the past data by itself. The Multi Layer Percep-tron Neural Network (MLPNN), Probabilistic Neural Network (PNN) Principal Component Analysis (PCA), Support Vector Machine (SVM) and Self Organizing Map (SOM) were also tested and benchmarked for their p
Highlights
Breast cancer is the most common tumor-related disease among women throughout the world, and the mortality rate caused by breast cancer is dramatically increasing
Six different neural network structure, Multi layer perceptron, Probabilistic neural network, Principal component analysis, Radial basis function, Support vector machine and Self organizing map neural network were applied to Wisconsin Breast Cancer Data (WBCD) database to show the performance of these neural networks on breast cancer data
The performance of various neural network structures was investigated for breast cancer diagnosis problem
Summary
Breast cancer is the most common tumor-related disease among women throughout the world, and the mortality rate caused by breast cancer is dramatically increasing. If the cancerous cells are detected before spreading to other organs, the survival rate for patient is more than 97% [25] which is the motivation factor to develop this automated diagnostic tool. Each system shows the PPV within the range from less than 60% (AdaBoost) up to over 95% (Neuro-Fuzzy Hybrid Models) Among those algorithms, Neuro-Fuzzy Hybrid models provide relatively remarkable performances in diagnosis. Neuro-Fuzzy Hybrid models provide relatively remarkable performances in diagnosis Those models are the combination of Neural Networks and Fuzzy Inference Systems encouraging the advantages and resolving the drawbacks of both NNs and FIS models
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
