Volume II | Issue 6 | 2026 | Paper ID: IJATEM23JUNE004

Teen Depression Level Classification Using a Crayfish Optimization Driven KAN-Residual Attention Network Based on Social Media Usage Patterns

Author(s)

, Pradheep T Rajan B

Abstract

Classifying teen depression levels using social media data is vital because it enables early detection, targeted interventions, and personalized mental health support for adolescents at risk. Therefore, in this paper, a hybrid Kolmogorov-Arnold Network (KAN)- Residual Attention Network (RAN) with Crayfish Optimization (CFO) is proposed for teen depression level classification based on social media usage. Data visualization in depression level classification by transforming complex behavioral, sensor, and psychological datasets, interpretable insights for machine learning models and clinicians. A KAN-RAN uses the feature-preserving nature of residual links and the contextual prioritization of attention mechanisms, this framework significantly enhances the detection of complex, non-linear biomarkers associated with depression. The CFO improves teen depression level classification by smart feature selector and hyperparameter tuner. The performance analysis is carried out using python software and assesses the model's performance using key metrics such as precision, recall, F1-score, confusion matrix, and ROC-AUC. The proposed KAN-RAN-CFO uses nine classes, they are relaxed, mild stress, low stress, moderate stress, elevated stress, high stress, severe stress, extreme stress and critical stress. The proposed KAN-RAN0CFO achieves high accuracy of 98% and better convergence speed. The proposed method outperforms conventional techniques, contributing a robust and reliable tool for teen depression level classification based on social media usage.

Keywords

KAN-RAN, CFO, Teen depression level classification, social media and accuracy.

References

[1] Rozi Karimi; Milad Baghalzadeh Shishehgarkhaneh; Robert C. Moehler; Yihai Fang; Saima Ahmad, Year: 2025, “Social Media Practices on Construction Sites: A New Conceptual Model of Social Media Impact on Team Performance”, Vol: 53, No: 6, pp. 147 – 167.

CrossRef Google Scholar Publisher Link
[2] Xinyi Wang; Loo-See Beh, Year: 2025, “Navigating New Terrain: Diverse Effects of Social Media on Employee Performance in China's Social Commerce Sector”, IEEE Transactions on Professional Communication, Vol: 68, No: 2, pp. 209 – 227.

CrossRef Google Scholar Publisher Link
[3] Thomas F. Stafford; B. Q. Duong, Year: 2023, “Social Media in Emerging Economies: A Cross-Cultural Comparison”, IEEE Transactions on Computational Social Systems, Vol: 10, No: 3, pp. 1160-1178.

CrossRef Google Scholar Publisher Link
[4] Praboda Rajapaksha; Reza Farahbakhsh, Year: 2019, “Scrutinizing News Media Cooperation in Facebook and Twitter”, IEEE Access, Vol: 13, pp. 123355-123368.

CrossRef Google Scholar Publisher Link
[5] Zixing Shen; S. Tan; M. J. Pritchard, Year: 2022, “Understanding the Effects of Visual Cueing on Social Media Engagement with YouTube Educational Videos”, IEEE Transactions on Professional Communication, Vol: 65, No: 2, pp. 337-350.

CrossRef Google Scholar Publisher Link
[6] Wei Wang; S. X. Zhang; Q. Liang; W. Deng, Year: 2024, “Social Media Levels the Playing Field of Entrepreneurial Entry for Disadvantaged Populations”, IEEE Transactions on Engineering Management, Vol: 71, pp. 9776-9789.

CrossRef Google Scholar Publisher Link
[7] Ujun Jeong; Bohan Jiang; Zhen Tan; H. Russell Bernard; Huan Liu, Year: 2024, “Descriptor: A Temporal Multi-network Dataset of Social Interactions in Bluesky Social (BlueTempNet)”, IEEE Data Descriptions, Vol: 1, pp. 71-79.

CrossRef Google Scholar Publisher Link
[8] Chen Zhang; Yue Wu; Ya Shen; Shengzhao Wang; Xuhui Zhu; Wei Shen, Year: 2024, “Adaptive Graph Convolutional Recurrent Network with Transformer and Whale Optimization Algorithm for Traffic Flow Prediction", Mathematics, Vol: 12, No: 10, pp. 1493.

CrossRef Google Scholar Publisher Link
[9] Wanlun Ma; Xiangyu Hu; Chao Chen; Sheng Wen, Year: 2022, “Social Media Event Prediction using DNN with Feedback Mechanism”, ACM Transactions on Management Information Systems, Vol: 13, No: 3, pp. 1-24.

CrossRef Google Scholar Publisher Link
[10] Ash Mohammad Abbas, Year: 2021, “Social network analysis using deep learning: applications and schemes”, Social Network Analysis and Mining, Vol: 11, No: 1, pp. 106.

CrossRef Google Scholar Publisher Link
[11] Jeya Bright Pankiraj; Govindaraj Vishnuvarthanan; Yudong Zhang; Murugan Pallikonda Rajasekaran; Ramaraj Kottamalai, Year: 2023, “Development of scalable coding on Encrypted Images using BTC for different Block Size”, Journal of Signal, Image and Video Processing, Vol: 17, No: 7, pp. 3821-3828.

CrossRef Google Scholar Publisher Link
[12] Kulandaivel Maruthamuthu Paramasivam; Prabakaran MP; Jeya Bright Pankiraj, Year: 2025, “CycleGAN algorithm based enhancement of night vision halation images from multiple sources”, AIP Advances, Vol: 15, No: 7, pp. 075031-1 - 075031-6.

CrossRef Google Scholar Publisher Link
[13] Dilip Kumar Sharma; Ningthoujam Chidananda Singh; Daneshwari A; Noola; Amala Nirmal Doss; Janaki Sivakumar, Year: 2025, “A review on various cryptographic techniques & algorithms”, Materials Today: Proceedings, Vol: 51, pp. 104-109.

CrossRef Google Scholar Publisher Link
[14] Thamilarasi V, Year: 2024, “Artificial intelligence-driven smart scenic management: Automated decision making and optimization”, Int. J. Intell. Syst. Appl. Eng, Vol: 12, pp. 2731.

Google Scholar

Full Length Article PDF

Article Details

Teen Depression Level Classification Using a Crayfish Optimization Driven KAN-Residual Attention Network Based on Social Media Usage Patterns

Author(s)

Abstract

Keywords

References

Quick Info