Abstract
Dyslexia is a neurological condition that presents difficulties and obstacles in learning, particularly in reading. Early diagnosis of dyslexia is crucial for children, as it allows the implementation of appropriate resources and specialized software to enhance their skills. However, the evaluation process can be expensive, time-consuming, and emotionally challenging. In recent years, researchers have turned to machine learning and deep learning techniques to detect dyslexia using datasets obtained from educational and healthcare institutions. Despite the existence of several deep learning models for dyslexia prediction, the problem of handling class imbalance significantly impacts the accuracy of detection. Therefore, this study proposes a robust deep learning model based on a variant of long short-term memory (LSTM) to address this issue. The advantage of Bidirectional LSTM, which has the ability to traverse both forward and backward, improves the pattern of understanding very effectively. Still, the problem of assigning values to the hyper-parameters in BLSTM is the toughest challenge which has to be assigned in a random manner. To overcome this difficulty, the proposed model induced a behavioral model known as Red Fox Optimization algorithm (RFO). Based on the inspiration of red fox searching behavior, this proposed work utilized the local and the global search in assigning and fine-tuning the values of hyper-parameters to handle the class imbalance in dyslexia dataset. The performance evaluation is conducted using two different dyslexia datasets (i.e., dyslexia 12_14 & real-time dataset). The simulation results explore that the proposed robust Bidirectional Long Short-Term Memory accomplishes the highest detection rate with reduced error rate compared to other deep learning models.
Similar content being viewed by others
Data availability
The dataset generated and analyzed during the current study are available from the corresponding author on reasonable request.
References
Heim S, Tschierse J, Amunts K, Wilms M, Vossel S, Willmes K, Grabowska A, Huber W. Cognitive subtypes of dyslexia. Acta Neurobiol Exp. 2008;68(1):73–82.
Wajuihian SO. Neurobiology of developmental dyslexia: part 1: a review of evidence from autopsy and structural neuro-imaging studies. South Afr Optometr. 2011;70(4):191–202.
Dowlin N, Gilad-Bachrach R, Laine K, Lauter K, Naehrig M, Wernsing J. Cryptonets: applying neural networks to encrypted data with high throughput and accuracy. In: 33rd International conference on machine learning; 2016, vol. 48. p. 342–351.
Pralhad GP, Joshi A, Chhipa M, Kumar S, Mishra G, Vishwakarma M. Dyslexia prediction using machine learning. In: 2021 International conference on artificial intelligence and machine vision (AIMV), Gandhinagar, India; 2021. p. 1–6.
Iwabuchi M, Hirabayashi R, Nakamura K, Dim NK. Machine learning based evaluation of reading and writing difficulties. Stud Health Technol Inf. 2017;242:1001.
Lai S, Xu L, Liu K, Zhao J. Recurrent convolutional neural networks for text classification. In: Proceedings of the twenty-ninth AAAI conference on artificial intelligence, Austin, TX, USA, 25–30 January 2015.
Kumar R, Srivastava S, Gupta J, Mohindru A. Diagonal recurrent neural network based identification of nonlinear dynamical systems with lyapunov stability based adaptive learning rates. Neurocomputing. 2018;287:102–17.
Gong L, Yu M, Jiang S, Cutsuridis V, Pearson S. Deep learning based prediction on greenhouse crop yield combined TCN and RNN. Sensors. 2021;21(13):4537.
Alex F, Larry MM. Features and machine learning for correlating and classifying between brain areas and dyslexia. CoRR abs/1812.10622; 2018.
JothiPrabha A, Bhargavi R, Harish B. Predictive model for dyslexia from eye fixation events. Int J Eng Adv Technol (IJEAT). 2019;9(1S3):235–40 (ISSN 2249-8958).
Mekyska J, Faundez-Zanuy M, Mzourek Z, Galaz Z, Smekal Z, Rosenblum S. Identification and rating of developmental dysgraphia by handwriting analysis. IEEE Trans Hum Mach Syst. 2017;47(2):235–48.
IzaSazanita I, Zahir MA, Ramlan SA, Li-Chih W, Sulaiman SN. CNN comparisons models on dyslexia handwriting classification. ESTEEM Acad J. 2021;17:12–25.
Usman OL, Muniyandi RC, Omar K, Mohamad M. Advance machine learning methods for dyslexia biomarker detection: a review of implementation details and challenges. IEEE Access. 2021;9:36879–97.
Chen A, Wijnen F, Koster C, Schnack H. Individualized early prediction of familial risk of dyslexia: a study of infant vocabulary development. Front Psychol. 2017;8:156.
Barraza J, Melin P, Valdez F, Gonzalez CI. fuzzy fireworks algorithm based on a sparks dispersion measure. Algorithms. 2017;10(3):83.
Rello L, Ballesteros M, Ali A, Serra M, Sánchez DA, Bigham JP. Dytective: diagnosing risk of dyslexia with a game, W4A. In: ’16 Proceedings of the 13th web for all conference, article no. 29, Montreal, Canada; April 11–13, 2016.
Geurts L, Vanden Abeele V, Celis V, Husson J, Van den Audenaeren L, Loyez L, Goeleven A, Wouters J, Ghesquière P. DIESEL-X: a game-based tool for early risk detection of dyslexia in preschoolers. In: Describing and studying domain-specific serious games. Springer International Publishing; 2015. p. 93–114.
Gaggi O, Palazzi CE, Ciman M, Galiazzo G, Franceschini S, Ruffino M, Gori S, Facoetti A. Serious games for early identification of developmental dyslexia. Comp Entertain. 2017;15(4):1–24.
Rauschenberger M, Rello L, Baeza-Yates R, Bigham JP. Towards language independent detection of dyslexia with a web-based game. W4A ‘18, Lyon, France, April 23–25, 2018.
Althelaya KA, El-Alfy E-SM, Mohammed S. Evaluation of bidirectional LSTM for short-and long-term stock market prediction. In: 2018 9th International conference on information and communication systems (ICICS), Irbid, Jordan; 2018. p. 151–156.
Al-Thelaya K, El-Alfy E-S, Mohammed S. Evaluation of bidirectional LSTM for short-and long-term stock market prediction; 2018. p. 151–156. https://doi.org/10.1109/IACS.2018.8355458.
Połap D, Wozniak M. Red fox optimization algorithm. Expert Syst Appl. 2021;166(15): 114107.
Funding
No funding received for this research.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of Interest
No conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the topical collection “Advances in Computational Approaches for Image Processing, Wireless Networks, Cloud Applications and Network Security” guest edited by P. Raviraj, Maode Ma and Roopashree H R.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zeema, J.L., Thirunavukkarasu, V., Sivabalan, R.V. et al. Robust Bidirectional Long Short-Term Memory-Based Class Imbalance Handling in Dyslexia Prediction at its Early Stage. SN COMPUT. SCI. 4, 605 (2023). https://doi.org/10.1007/s42979-023-02049-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s42979-023-02049-9