Optimal Predictive Model for FIFA World Cup
Durgansh Sharma1, Vaibhava Sharma2

1Durgansh Sharma, Associate Professor, School of Computer Science, University of Petroleum and Energy Studies, Dehradun (Uttarakhand), India.
2Vaibhava Sharma, Accounts Executive, Finance Team, Next Level Business Services Inc., Noida (Uttar Pradesh), India.
Manuscript received on 27 March 2019 | Revised Manuscript received on 04 April 2019 | Manuscript Published on 12 April 2019 | PP: 118-123 | Volume-7 Issue-6C April 2019 | Retrieval Number: F90410476C19/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: In this research, the proposed deep learning network uses H2O framework using Multi-layer Feed Forward Network. Statistics of the 592 FIFA world cup Matches collected and used to train Naïve Bayes, k-NN and deep learning networks. As observed, the efficiency of deep learning based network is superior as compared with Naïve Bayes and k-NN for prediction of sports especially football/ soccer using dataset of Match result and status. The prediction observed with 97.68% accuracy while keeping the training-testing ratio as 20:80 using Deep Learning using “Max-out with Dropout” activation function. The model k-NN and Naïve Bayes were trained with 80:20 training-testing ratio.
Keywords: Predictive Analytics, Deep Learning, Naïve Bayes, k-NN, FIFA.
Scope of the Article: Predictive Analysis