Call for Paper - November 2022 Edition
IJCA solicits original research papers for the November 2022 Edition. Last date of manuscript submission is October 20, 2022. Read More

New Approach for Joint Multilabel Classification with Community-Aware Label Graph Learning Technique

International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
Year of Publication: 2018
R. V. Argiddi, Disha Rajan Shah

R V Argiddi and Disha Rajan Shah. New Approach for Joint Multilabel Classification with Community-Aware Label Graph Learning Technique. International Journal of Computer Applications 180(36):1-7, April 2018. BibTeX

	author = {R. V. Argiddi and Disha Rajan Shah},
	title = {New Approach for Joint Multilabel Classification with Community-Aware Label Graph Learning Technique},
	journal = {International Journal of Computer Applications},
	issue_date = {April 2018},
	volume = {180},
	number = {36},
	month = {Apr},
	year = {2018},
	issn = {0975-8887},
	pages = {1-7},
	numpages = {7},
	url = {},
	doi = {10.5120/ijca2018916892},
	publisher = {Foundation of Computer Science (FCS), NY, USA},
	address = {New York, USA}


Multi-label classification is a significant machine learning task in which one allocates a subset of candidate labels to an object. A new multi-label classification technique based on Conditional Bernoulli Mixtures. Exploiting label dependency for multi-label image classification cans considerably develop classification performance. Probabilistic Graphical Models are one of the primary methods for demonstrating such dependences. The structure of graphical models, however, is which ever resolute heuristically or learned from very inadequate information. Moreover, neither of these methodologies scales well to large or complex graphs. We recommend a principled way to learn the structure of a graphical model by in view of input features and labels, composed with loss functions. We formulate this problem into a max-margin framework primarily, and then convert it into a convex programming problem. In conclusion, we suggest a highly scalable technique that activates a set of cliques iteratively. Our methodology exhibits both strong theoretical properties and a substantial performance development over state-of-the-art approaches on both synthetic and real-world data sets. Our proposed system has numerous attractive properties: it captures label dependences; it decreases the multi-label problem to numerous standard binary and multi-class problems; it subsumes the classic independent binary prediction and power-set subset prediction approaches as special cases; and it exhibitions accuracy and/or computational complexity benefits over present approaches. We demonstrate two implementations of our technique by means of logistic regressions and gradient boosted trees, organized with a simple training procedure centered on Expectation Maximization. We promote derive an efficient prediction procedure centered on dynamic programming, thus avoiding the cost of scrutinizing an exponential number of probable label subsets. For the testing we will use and show the efficiency of the proposed method in contradiction of competitive substitutes on benchmark datasets with image as well as pdf.


  1. X. Kong, M. K. Ng, and Z.-H. Zhou, “Transductivemultilabel learning via label set propagation,” IEEE Trans. Knowl. Data Eng., vol. 25, no. 3, pp. 704–719, Mar. 2013.
  2. J. Read, B. Pfahringer, G. Holmes, and E. Frank, “Classifier chains for multi-label classification,” J. Mach. Learn., vol. 85, no. 3, pp. 333–359, Dec. 2011.
  3. M.-L. Zhang and Z.-H.Zhou, “Multilabel neural networks with applications to functional genomics and text categorization,” IEEE Trans. Knowl. Data Eng., vol. 18, no. 10, pp. 1338–1351, Oct. 2006.
  4. G. Tsoumakas, I. Katakis, and L. Vlahavas, “Random k-label sets for multilabel classification,” IEEE Trans. Knowl. Data Eng., vol. 23, no. 7, pp. 1079–1089, Jul. 2011.
  5. Ant Colony optimization L Jiao, L Feng - Information and Computing (ICIC), 2010 -
  6. A Survey on Approaches of Multirelational Classification Based On Relational Database ShraddhaModi, AmitThakkar, AmitGanatra, International Journal of Engineering and Advanced Technology (IJEAT) ISSN: 2249 – 8958, Volume-1, Issue-3
  7. Lachetar, N. ;Comput. Sci. Dept., Univ. 20 Aout 1955, Skikda, Algeria ; Bahi, H. Application of an ant colony algorithm for text indexing, :Multimedia Computing and Systems (ICMCS), 2011 International Conference –IEEE 2011.


Multi graph learning, Document classification approach, semi supervised learning.