IJSTR

International Journal of Scientific & Technology Research

Home About Us Scope Editorial Board Blog/Latest News Contact Us
0.2
2019CiteScore
 
10th percentile
Powered by  Scopus
Scopus coverage:
Nov 2018 to May 2020

CALL FOR PAPERS
AUTHORS
DOWNLOADS
CONTACT

IJSTR >> Volume 9 - Issue 4, April 2020 Edition



International Journal of Scientific & Technology Research  
International Journal of Scientific & Technology Research

Website: http://www.ijstr.org

ISSN 2277-8616



An Optimal Deep Neural Network Model for Lymph Disease Identification And Classification

[Full Text]

 

AUTHOR(S)

J. Junia Deborah, Dr. Latha Parthiban

 

KEYWORDS

Lymph disease, Classification, Deep Learning, Adam Optimizer

 

ABSTRACT

At present days, deep learning (DL) models find useful in various domains especially healthcare sector. This paper introduces a new DL based lymph disease identification and classification model. An optimal deep neural network (DNN) model is applied to classify the lymph data utilizing the stacked autoencoders (SA) which is generally used to extract the features from the dataset which is further classified by the utilization of SoftMax layer. In addition, to make the DNN model more efficient, DL based Adam optimizer (AO) is used which is an adaptive learning rate optimization algorithm which has been developed particularly to train DNN. The presented DNN with AO model called DNN-AO model is tested against benchmark Lymphography dataset and the results are assessed under diverse measures. The experimental outcome verified that effective classification performance is showcased by the DNN-AO model with the maximum sensitivity, specificity, accuracy and kappa value of 95.59, 95.95, 95.95 and 91.69 respectively.

 

REFERENCES

[1] K.J. Ciosa, G.W. Mooree, Uniqueness of medical data mining,Artif. Intell. Med. 26 (2002) 1–24.
[2] W. Ceusters, Medical natural language understanding as a supporting technology for data mining in healthcare, in: K.J.Cios (Ed.), Medical Data Mining and Knowledge Discovery,Springer, Heidelberg, 2000, pp. 32–60 (Chapter 3).
[3] I. Kononenko, Machine learning for medical diagnosis:history state of the art and perspective, Artif. Intell. Med. 23(2001) 89–109.
[4] F. Calle-Alonso, C.J. Pérez, J.P. Arias-Nicolás, J. Martín,Computer-aided diagnosis system: a Bayesian hybridclassification method, Comput. Methods Programs Biomed.112 (2013) 104–113.
[5] S.H. Huang, L.R. Wulsin, H. Li, J. Guo, Dimensionalityreduction for knowledge discovery in medical claimsdatabase: application to antidepressant medicationutilization study, Comput. Methods Programs Biomed. 93(2009) 115–123.
[6] Z. Cselényi, Mapping the dimensionality density andtopology of data: the growing adaptive neural gas, Comput.Methods Programs Biomed. 78 (2005) 141–156.
[7] I.A. Gheyas, L.S. Smith, Feature subset selection in largedimensionality domains, Pattern Recognit. 43 (2010)5–13.
[8] H.H. Inbarani, A.T. Azar, G. Jothi, Supervised hybrid featureselection based on PSO and rough sets for medicaldiagnosis, Comput. Methods Programs Biomed. (2013),Available online 16 October 2013, ISSN 0169-2607,http://dx.doi.org/10.1016/j.cmpb.2013.10.007
[9] M. Macaˇs, L. Lhotská, E. Bakstein, D. Novák, J. Wild, T. Sieger,P.Vostatek, R. Jech, Wrapper feature selection for smallsample size data driven by complete error estimates,Comput. Methods Programs Biomed. 108 (2012) 138–150.
[10] CancerResearchUK.http://www.cancerresearchuk.org(accessed 03.11.13).
[11] A. Luciani, E. Itti, A. Rahmouni, Lymph node imaging: basicprinciples, Eur. J. Radiol. 58 (2006) 338–344.
[12] N. Jahan, P. Narayanan, A. Rockall, Magnetic resonancelymphography in gynaecological malignancies, CancerImaging 10 (2010) 85–96.
[13] R. Sharma, J.A. Wendt, J.C. Rasmussen, A.E. Adams, M.V.Marshall, E.M. Sevick-Muraca., New horizons for imaginglymphatic function, Ann. N. Y. Acad. Sci. 1131 (2008)13–36.
[14] A. Guermazi, P. Brice, C. Hennequin, E. Sarfati,Lymphography: an old technique retains its usefulness,Radiographics 23 (2003) 1541–1558.
[15] K. Polat, S. Gunes, A novel hybrid intelligent method basedon C4.5 decision tree classifier and one-against-all approachfor multi-class classification problems, Expert Syst. Appl.:Int. J. 36 (2009) 1587–1592.
[16] G. Iannello, G. Percannella, C. Sansone, P. Soda, On the use ofclassification reliability for improving performance of theone-per-class decomposition method, Data Knowl. Eng. 68(2009) 1398–1410.
[17] Kannadasan, K., Edla, D.R. and Kuppili, V., 2018. Type 2 diabetes data classification using stacked autoencoders in deep neural networks. Clinical Epidemiology and Global Health.
[18] UCI.MachineLearningRepository.http://archive.ics.uci.edu/ml/index.html (accessed 03.11.13).