Efficient Technique for Cancer Prediction Using ANN Classification Based on pH Parameter

By considering the growth of population in the present situation, there should be a high level and very effective healthcare analysis is the basic criteria. It should be the same situation at both home and hospitals. There are plenty of researches are done on human health level condition, research in data mining for health analysis is most significant in order to provide good services to the patients health. These researches are becoming a significant opportunity for improving the quality of health care services. In this regard, this paper presents an innovative approach for human health analysis using pH value of blood and urine samples. Now a day it is difficult for the doctor to detect the hypercalcaemia condition which may lead to different cancer disease. So to analysis the patient electrolyte value the doctor come to know that the volume of the electrolytes and made the judgment whether the person is normal or not. To analyze the patient health the blood and urine reports of the patients with different conditions are gathered as input parameters. Both blood and urine values are fused together by concatenation data fusion technique to form as single data. Dimension reduction is also implemented to convert the high dimensional data samples into the low dimensional space, so that the intensive information contained in the data is protected. As the dimensionality of data gets reduced, it encourage improving the robustness of the classifier and decreases computational complexity. In this paper, reduction is carried out by Principal Component Analysis on fused data. Based on feature level and on the basis of PCA technique Feature extraction is applied to extract important features from considered data. For classification ANN classifier is implemented as it provides best computational speed and accuracy. In accordance with the classification outcome overall health analysis results are estimated.