A Deep Learning Model for Classification of Cancer Types on Gene Expression Data

Main Article Content

P. Avila Clemenshia
C. Deepa


Feature selection, Local Optimum, Dimensionality reduction, weight values and Fuzzy Logic


Cancer Classification have great significance in cancer diagnostics and inventions of new drugs. Earlier researches in this area mostly focused on clinical aspects with low diagnostic capabilities. Classifying cancers using gene expression data have the ability to address most preliminary issues associated with diagnosis of cancers or inventions of drugs. Advancements in DNA micro-array approaches have opened the way to monitor thousands of gene expressions. This enterprising quality of gene expression data has been the impetus of this study which examines the feasibility of identifying cancers from gene expression data. For analysis of tumor types, this work proposed DFN Forest (Deep Flexible Neural Forest) model and introduced an improved model for cancer types classification. In this work, Principal Component Analysis (PCA) algorithm is used for dimensionality reduction. Feature selection is done with the help of ICA (Imperialist Competitive Algorithm). DFFN Forest (Deep Fuzzy Flexible Neural Forest), which uses fuzzy logic to update the weight values, is used in this work to classify cancer subtypes. Results from experiments show that the proposed model is effective in terms of metrics like precision, recall, accuracy, and error rate.

Abstract 64 | PDF Downloads 99


1. Nidheesh, N., Nazeer, K. A., & Ameer, P. M. (2017). An enhanced deterministic K-Means clusteringalgorithmforcancersubtypepredictionfromgeneexpressiondata. Computersinbiologyandmedicine,91,213-221.
2. Ujjwal Maulik, Anirban Mukhopadhyay andDebasis Chakraborty, “Gene-Expression-Based CancerSubtypesPredictionThroughFeatureSelectionandTransductiveSVM”,Vol.60,issues(4),2013
3. De Kruijf, E. M., Engels, C. C., van de Water, W., Bastiaannet, E., Smit, V. T., van de Velde, C. J., ... &Kuppen, P. J. (2013). Tumor immune subtypes distinguish tumor subclasses with clinical implications inbreastcancer patients.Breastcancerresearchand treatment,142(2), 355-364.
4. Prat,Aleix,EstelaPineda,BarbaraAdamo,PatriciaGalván,AranzazuFernández,LydiaGaba,MarcDíez,Margarita Viladot, Ana Arance, and Montserrat Muñoz. "Clinical implications of the intrinsic
molecularsubtypesof breastcancer."TheBreast24 (2015): S26-S35.
5. Thanki,K.,Nicholls,M.E.,Gajjar,A.,Senagore,A.J.,Qiu,S.,Szabo,C.,...&Chao,C.(2017).Consensusmolecularsubtypesofcolorectalcancerandtheirclinicalimplications. Internationalbiologicalandbiomedicaljournal,3(3),105.
6. Tomczak, K., Czerwińska, P., & Wiznerowicz,M. (2015). The Cancer Genome Atlas (TCGA): animmeasurablesourceofknowledge.Contemporary oncology,19(1A), A68.
7. Finnegan, Timothy J., and Lisa A. Carey. "Gene-expression analysis and the basal-like breast cancersubtype." (2007): pp.55-63.
8. Teschendorff, A. E., Miremadi, A., Pinder, S. E., Ellis, I. O., & Caldas, C. (2007). An immune responsegeneexpressionmoduleidentifiesagoodprognosissubtypeinestrogenreceptornegativebreastcancer. Genomebiology, 8(8), R157.
9. Wong,G.,Leckie,C.,&Kowalczyk,A.(2011).FSR:featuresetreductionforscalableandaccuratemulti-classcancer subtype classificationbased oncopynumber.Bioinformatics, 28(2),pp.151-159.
10. Zhang, W., Feng, H., Wu, H., & Zheng, X. (2017). Accounting for tumor purity improves cancer subtypeclassificationfrom DNAmethylation data.Bioinformatics, 33(17),2651-2657.
11. Gao,Yuan,andGeorgeChurch."Improvingmolecularcancerclassdiscoverythroughsparsenon-negativematrixfactorization."Bioinformatics 21, no. 21 (2005): pp.3970-3975.
12. Jinn-Yi Yeh, Tai-Shi Wu, Min-Che Wu, Der-Ming Chang, Applying Data Mining Techniques for CancerClassification from Gene Expression Data, IEEE International Conference on Convergence InformationTechnology,2007
13. Atashpaz-Gargari, E.; Lucas, C. Imperialist competitive algorithm for minimum bit error rate beamforming. Int. J. Bio-Inspired Comput. 2009, 1, 125–133. [Google Scholar]
14. Rasul, E.; JavedaniSadaei, H.; Abdullah, A.H.; Gani, A. Imperialist competitive algorithm combined with refined high-order weighted fuzzy time series (RHWFTS–ICA) for short term load forecasting. Energy Convers. Manag. 2013, 76, 1104–1116. [Google Scholar]
15. Shamshirband, S.; Amini, A.; Nor Badrul, A.; Mat Kiah, M.L.; Ying, W.T.; Furnell, S. D-FICCA: A density-based fuzzy imperialist competitive clustering algorithm for intrusion detection in wireless sensor networks. J. Int. Meas. Confed. 2014, 55, 212–226. [Google Scholar] [CrossRef]
16. Atashpaz-Gargari, E.; Lucas, C. Imperialist competitive algorithm: An algorithm for optimization inspired by imperialistic competition. In Proceedings of the IEEE Congress on Evolutionary Computation (CEC 2007), Singapore, 25–28 September 2007; pp. 4661–4667.
17. Nourmohammadia, A.; Zandiehb, M.; Tavakkoli-Moghaddamca, R. An imperialist competitive algorithm for multi-objective U-type assembly line design. J. Comput. Sci. 2012, 4, 393–400. [Google Scholar] [CrossRef]
18. Banaei, M.; Seyed-Shenava, S.; Farahbakhsh, P. Dynamic stability enhancement of power system based on a typical unified power flow controllers using imperialist competitive algorithm. Ain Shams Eng. J. 2014, 5, 691–702. [Google Scholar] [CrossRef]
19. Hadidi, A.; Hadidi, M.; Nazari, A. A new design approach for shell-and-tube heat exchangers using imperialist competitive algorithm (ICA) from economic point of view. Energy Convers. Manag. 2013,