Sage Journals: Discover world-class research

Abstract

Esports is a rapidly growing industry with a massive viewership, generating a wealth of data that can be leveraged to improve win prediction. Machine learning approaches have been employed for game outcome prediction. However, they have typically focused on a single classifier/predictor, which exhibits effectiveness and scalability issues due to intricate architectures and a lack of parallel processing. This paper proposes a new hybrid parallel ensemble learning framework combined with the Fuzzy ARTMAP (FAM) model to enhance win prediction in professional esports. The proposed framework leverages parallel computing and combines diverse deep neural network (DNN) and machine learning (ML) models as base learners, with FAM serving as the meta-learner to effectively integrate and train the base learners outputs. Empirical studies using a c demonstrate the superior efficacy and efficiency of this approach compared to conventional methods. The results show that HELP-GOP is effective and efficient, achieving a high-performance level with 98.49% accuracy and a speedup of up to 5.2× with parallel processing. Additionally, the results demonstrate that hybrid parallel ensemble learning represents a significant advancement in win prediction for professional esports, showcasing more accurate and sophisticated predictive analytics in the esports field.

Keywords

esports ensemble learning machine learning fuzzy ARTMAP deep neural network parallel computing

Get full access to this article

View all access options for this article.

References

Agarwala

Pearce

(2014). Learning Dota 2 team compositions. Sl: sn.

Aha

D. W.

Kibler

Albert

M. K.

(1991). Instance-based learning algorithms. Machine Learning, 6(1991), 37–66. https://doi.org/10.1007/BF00153759

Ahn

Collis

Jenny

(2020). The one billion dollar myth: Methods for sizing the massively undervalued esports revenue landscape. International Journal of Esports, 1(1). https://www.ijesports.org/article/15/html

Akhmedov

Phan

A. H.

(2021). Machine learning models for DOTA 2 outcomes prediction. arXiv preprint arXiv:2106.01782.

Al-Andoli

M. N.

Sim

K. S.

Tan

S. C.

Goh

P. Y.

Lim

C. P.

(2023). An ensemble-based parallel deep learning classifier with PSO-BP optimization for malware detection. IEEE Access, 11(2023), 76330–76346. https://doi.org/10.1109/ACCESS.2023.3296789

Anshori

Mar’i

Alauddin

M. W.

Bachtiar

F. A.

(2018). Prediction result of dota 2 games using improved svm classifier based on particle swarm optimization. IEEE, pp.121–126.

Berner

, et al. (2019). Dota 2 with large scale deep reinforcement learning. arXiv preprint arXiv:1912.06680.

Block

Haack

(2021). eSports: a new industry. EDP Sciences, p.04002.

Breiman

(2001). Random forest, vol. 45. Mach Learn, vol. 1.

10.

Brown

(2024). Rethinking remakes: Value and culture in video game temporalization. Games and Culture, 19(3), 337–356. https://doi.org/10.1177/15554120231163655

11.

Carpenter

G. A.

Grossberg

Markuzon

Reynolds

J. H.

Rosen

D. B.

(1992). Fuzzy ARTMAP: A neural network architecture for incremental supervised learning of analog multidimensional maps. IEEE Transactions on Neural Networks, 3(5), 698–713. https://doi.org/10.1109/72.159059

12.

Casselman

(2024). Resistance is futile: eSports is massive… and growing. ESPN. https://www.espn.com/espn/story/_/id/13059210/esports-massive-industry-growing (accessed 2024)

13.

Dallmann

Kohlmann

Zoller

Hotho

(2021). Sequential item recommendation in the moba game dota 2. IEEE, pp.10–17.

14.

Formosa

, et al. (2022). Definitions of esports: A systematic review and thematic analysis. Proceedings of the ACM on Human-Computer Interaction, 6(CHI PLAY), 1–45. https://doi.org/10.1145/3549490

15.

Gough

(2022). eSports audience size worldwide from 2020 to 2025. Statista. com.

16.

Guo

Chen

Hong

(2022). Pattern recognition and analysis: neural network using weighted cross entropy (1 ed.). IOP Publishing, p.012043.

17.

Haykin

(2009). Neural networks and learning machines, 3/E. Pearson Education India.

18.

Hochreiter

Schmidhuber

(1997). Long short-term memory. Neural Computation, 9(8), 1735–1780. https://doi.org/10.1162/neco.1997.9.8.1735

19.

Hodge

V. J.

Devlin

Sephton

Block

Cowling

P. I.

Drachen

(2019). Win prediction in multiplayer esports: Live professional match prediction. IEEE Transactions on Games, 13(4), 368–379. https://doi.org/10.1109/TG.2019.2948469

20.

Jin

(2013). Fuzzy ARTMAP ensemble based decision making and application. Mathematical Problems in Engineering, 2013(1), 124263. https://doi.org/10.1155/2013/124263

21.

Joe

(2017). Dota 2 Matches Dataset. Kaggle.

22.

John

G. H.

Langley

(2013). Estimating continuous distributions in Bayesian classifiers. arXiv preprint arXiv:1302.4964.

23.

Katona

, et al. (2019). Time to die: Death prediction in dota 2 using deep learning. IEEE, pp.1–8.

24.

C. H.

, et al. (2022). DOTA 2 match prediction through deep learning team fight models. IEEE, pp.96–103.

25.

Kinkade

Lim

(2015). Dota 2 win prediction. University of California, 1(2015), 1–13. http://jmcauley.ucsd.edu/cse258/projects/fa15/018.pdf

26.

Looi

Dhaliwal

Alhajj

Rokne

(2018). Recommender system for items in dota 2. IEEE Transactions on Games, 11(4), 396–404. https://doi.org/10.1109/TG.2018.2844121

27.

Makarov

Savostyanov

Litvyakov

Ignatov

D. I.

(2018). Predicting winning team and probabilistic ratings in “Dota 2” and “Counter-Strike: Global Offensive” video games. Springer, pp.183–196.

28.

Mohammed

K. N. A.-A.

Swee Sim

S. C. T.

Seera

Lim

C. P.

(2023). A parallel ensemble learning model for fault detection and diagnosis of industrial machinery. IEEE Access, 11(2023), 39866–39878. https://doi.org/10.1109/ACCESS.2023.3267089

29.

Mohammed Nasser Al-Andoli

S. C. T.

Sim

K. S.

Goh

P. Y.

Lim

C. P.

(2023). An ensemble deep learning classifier stacked with fuzzy ARTMAP for malware detection. Journal of Intelligent and Fuzzy Systems, 44(6), 10477–10493. https://doi.org/10.3233/JIFS-230009

30.

Mora-Cantallops

Sicilia

M-Á

(2018). MOBA Games: A literature review. Entertainment Computing, 26(1), 128–138. https://doi.org/10.1016/j.entcom.2018.02.005

31.

Moritz

, et al. (2018). Ray: A distributed framework for emerging {AI} applications. In 13th {USENIX} symposium on operating systems design and implementation ({OSDI} 18) (pp. 561–577).

32.

Musika

Malathong

Arthayukti

Konpang

Jarernpon

Thitimahatthanakusol

(2024). Predicting Dota 2 Game Period Winners Through Hero Analysis. IEEE, pp.01–05.

33.

Norouzzadeh

Y. R.

Spronck

Sifa

Drachen

(2017). Predicting Victory in a Hybrid Online Competitive Game: The Case of Destiny. York, pp.207–214.

34.

Pepatang

Kotrajaras

(2024). Machine Learning Model for Skill Level Prediction in Multiplayer Online Battle Arena. IEEE, pp.1–6.

35.

Putra

F. A.

Qomariyah

N. N.

(2022). Dota 2 win loss prediction from item and hero data with machine learning. IEEE, pp.204–209.

36.

Schubert

Drachen

Mahlmann

(2016). Esports analytics through encounter detection other sports, ed.

37.

Song

Zhang

(2015). Predicting the winning side of DotA2. Sl: sn.

38.

Tao

Yang

Benslimane

Liu

(2024). Predicting Dota 2 Game Outcomes Using Logistic Regression and Decision Tree Models. IEEE, pp.187–191.

39.

Uddin

Fahmida

Moyeen

S. I.

Hasan

M. M.

(2022). DOTA2 Winner Team Prediction based on Stacked Bidirectional LSTM Network. IEEE, pp.1–5.

40.

Werbos

P. J.

(1990). Backpropagation through time: What it does and how to do it. Proceedings of the IEEE, 78(10), 1550–1560. https://doi.org/10.1109/5.58337

41.

Wang

Xuan

(2016). A selective fuzzy ARTMAP ensemble and its application to the fault diagnosis of rolling element bearing. Neurocomputing, 182(2016), 25–35. https://doi.org/10.1016/j.neucom.2015.12.015

42.

Zhang

Chen

Xie

(2019). Moba-slice: A time slice based evaluation framework of relative advantage between teams in moba games. Springer, pp.23–40.

43.

Zhang

Gao

Han

Tian

(2020). Improved Dota2 lineup recommendation model based on a bidirectional LSTM. Tsinghua Science and Technology, 25(6), 712–720. https://doi.org/10.26599/TST.2019.9010065

44.

Zhou

(2024). Emerging markets in global esports: Opportunities and challenges.

A Hybrid Parallel Ensemble Learning Framework with Fuzzy ARTMAP to Improve Outcome Prediction in Professional Esports

Abstract

Keywords

Get full access to this article

References