# Student Behavior Prediction of Mental Health Based on Two-Stream Informer Network

^{*}

## Abstract

**:**

## 1. Introduction

- An architecture based on a two-stream informer is designed. The Time Encoder and Behavior Encoder are used to capture the interdependence between students’ behaviors and the time cycle trend, respectively;
- In order to prevent information loss, the features of the two channels are fused using an Intermediate Fusion Module;
- A dataset with 8213 students’ behavioral data is made for experimental analysis; the TSIN is evaluated on this dataset and 11 multivariate time-series benchmark datasets, and a comprehensive ablation study is conducted with other advanced deep learning models. The experiments show that the TSIN has good performance.

## 2. Related Work

#### 2.1. Education Data Mining

#### 2.2. Time-Series Classification

## 3. Proposed Method

#### 3.1. Problem Description

#### 3.2. Overview of the Framework

#### 3.3. Input Layer

#### 3.4. Encoding Layer

#### 3.4.1. Time Encoder

#### 3.4.2. Behavior Encoder

#### 3.5. Intermediate Fusion Module

#### 3.6. Classification Prediction Module

## 4. Experimental Results and Analysis

#### 4.1. Basic Settings

#### 4.1.1. Dataset

#### 4.1.2. Comparison Methods

- Universal Neural Network coder (Encoder) [38].
- Multi-scale Convolutional Neural Network (MCNN) [28].
- Multi-Channel Deep Convolutional Neural Network (MCDCNN) [29].
- Time Convolutional Neural Network (Time-CNN) [39].
- Time Le-Net (t-LeNet) [40].
- Time Warping Invariant Echo State Network (TWIESN) [30].
- Gated Transformer Network (GTN) for multivariate time-series classification [32].

#### 4.1.3. Parameter Setup and Experimental Support

#### 4.1.4. Evaluation Metric

#### 4.2. Comparison with Representative Works

#### 4.2.1. Accuracy Comparison

#### 4.2.2. Time Complexity Comparison

#### 4.3. Ablation Test

#### 4.4. Analysis of Attention Map

#### 4.5. Analysis of Features

## 5. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## References and Notes

- Chu, Y.; Yin, X. Data Analysis of College Students’ Mental Health Based on Clustering Analysis Algorithm. Complexity
**2021**, 2021, 9996146. [Google Scholar] [CrossRef] - Tang, Q.; Zhao, Y.; Wei, Y.; Jiang, L. Research on the mental health of college students based on fuzzy clustering algorithm. Secur. Commun. Netw.
**2021**, 2021, 3960559. [Google Scholar] [CrossRef] - Baker, R. Data mining for education. Int. Encycl. Educ.
**2010**, 7, 112–118. [Google Scholar] - Hsieh, K.Y.; Hsiao, R.C.; Yang, Y.H.; Lee, K.H.; Yen, C.F. Relationship between self-identity confusion and internet addiction among college students: The mediating effects of psychological inflexibility and experiential avoidance. Int. J. Environ. Res. Public Health
**2019**, 16, 3225. [Google Scholar] [CrossRef] [PubMed] - Ding, Y.; Chen, X.; Fu, Q.; Zhong, S. A depression recognition method for college students using deep integrated support vector algorithm. IEEE Access
**2020**, 8, 75616–75629. [Google Scholar] [CrossRef] - Akram, A.; Fu, C.; Li, Y.; Javed, M.Y.; Lin, R.; Jiang, Y.; Tang, Y. Predicting students’ academic procrastination in blended learning course using homework submission data. IEEE Access
**2019**, 7, 102487–102498. [Google Scholar] [CrossRef] - Yang, Z.; Su, Z.; Liu, S.; Liu, Z.; Ke, W.; Zhao, L. Evolution features and behavior characters of friendship networks on campus life. Expert Syst. Appl.
**2020**, 158, 113519. [Google Scholar] [CrossRef] - Wang, Y.; Wang, Q.W.; Tao, Y.Y.; Xie, W.W. Empirical Study of Consumption Behavior of College Students under the Influence of Internet-based Financing Services. Procedia Comput. Sci.
**2021**, 187, 152–157. [Google Scholar] [CrossRef] - Lim, H.; Kim, S.; Chung, K.M.; Lee, K.; Kim, T.; Heo, J. Is college students’ trajectory associated with academic performance? Comput. Educ.
**2022**, 178, 104397. [Google Scholar] [CrossRef] - Asif, R.; Merceron, A.; Ali, S.A.; Haider, N.G. Analyzing undergraduate students’ performance using educational data mining. Comput. Educ.
**2017**, 113, 177–194. [Google Scholar] [CrossRef] - Su, Y.; Liu, Q.; Liu, Q.; Huang, Z.; Yin, Y.; Chen, E.; Ding, C.; Wei, S.; Hu, G. Exercise-enhanced sequential modeling for student performance prediction. In Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, LA, USA, 2–7 February 2018; Volume 32. [Google Scholar] [CrossRef]
- Wu, F.; Zheng, Q.; Tian, F.; Suo, Z.; Zhou, Y.; Chao, K.M.; Xu, M.; Shah, N.; Liu, J.; Li, F. Supporting poverty-stricken college students in smart campus. Future Gener. Comput. Syst.
**2020**, 111, 599–616. [Google Scholar] [CrossRef] - Ma, Y.; Zhang, X.; Di, X.; Ren, T.; Yang, H.; Cai, B. Analysis and identification of students with financial difficulties: A behavioural feature perspective. Discret. Dyn. Nat. Soc.
**2020**, 2020, 071025. [Google Scholar] [CrossRef] - Yao, H.; Lian, D.; Cao, Y.; Wu, Y.; Zhou, T. Predicting Academic Performance for College Students: A Campus Behavior Perspective. ACM Trans. Intell. Syst. Technol.
**2019**, 10, 1–21. [Google Scholar] [CrossRef] - Govindasamya, K.; Velmuruganb, T. A study on classification and clustering data mining algorithms based on students academic performance prediction. Int. J. Control. Theory Appl.
**2017**, 10, 147–160. [Google Scholar] - Li, Y.; Zhang, H.; Liu, S. Applying data mining techniques with data of campus card system. In Proceedings of the IOP Conference Series: Materials Science and Engineering, Ulaanbaatar, Mongolia, 10–13 September 2020; Volume 715, p. 012021. [Google Scholar] [CrossRef]
- Morelli, S.A.; Ong, D.C.; Makati, R.; Jackson, M.O.; Zaki, J. Empathy and well-being correlate with centrality in different social networks. Proc. Natl. Acad. Sci. USA
**2017**, 114, 9843–9847. [Google Scholar] [CrossRef] [PubMed] - Ding, D.; Li, J.; Wang, H.; Liang, Z. Student behavior clustering method based on campus big data. In Proceedings of the 2017 13th International Conference on Computational Intelligence and Security (CIS), Hong Kong, China, 15–18 December 2017; pp. 500–503. [Google Scholar] [CrossRef]
- Chen, M.; Jiang, S. Analysis and research on mental health of college students based on cognitive computing. Cogn. Syst. Res.
**2019**, 56, 151–158. [Google Scholar] [CrossRef] - Pedrelli, P.; Nyer, M.; Yeung, A.; Zulauf, C.; Wilens, T. College students: Mental health problems and treatment considerations. Acad. Psychiatry
**2015**, 39, 503–511. [Google Scholar] [CrossRef] - Hokanson, J.E.; Rubert, M.P.; Welker, R.A.; Hollander, G.R.; Hedeen, C. Interpersonal concomitants and antecedents of depression among college students. J. Abnorm. Psychol.
**1989**, 98, 209. [Google Scholar] [CrossRef] - Abanda, A.; Mori, U.; Lozano, J.A. A review on distance based time series classification. Data Min. Knowl. Discov.
**2019**, 33, 378–412. [Google Scholar] [CrossRef] - Ismail, F.H.; Forestier, G.; Weber, J.; Idoumghar, L.; Muller, P.A. Deep learning for time series classification: A review. Data Min. Knowl. Discov.
**2019**, 33, 917–963. [Google Scholar] [CrossRef] - Jalalian, A.; Chalup, S.K. GDTW-P-SVMs: Variable-length time series analysis using support vector machines. Neurocomputing
**2013**, 99, 270–282. [Google Scholar] [CrossRef] - Yamada, Y.; Suzuki, E.; Yokoi, H.; Takabayashi, K. Decision-tree induction from time-series data based on a standard-example split test. In Proceedings of the 20th international conference on machine learning (ICML-03), Washington, DC, USA, 21–24 August 2003; pp. 840–847. [Google Scholar]
- Gupta, A.; Gupta, H.P.; Biswas, B.; Dutta, T. An early classification approach for multivariate time series of on-vehicle sensors in transportation. IEEE Trans. Intell. Transp. Syst.
**2020**, 21, 5316–5327. [Google Scholar] [CrossRef] - Wang, Z.; Yan, W.; Oates, T. Time series classification from scratch with deep neural networks: A strong baseline. In Proceedings of the 2017 International joint conference on neural networks (IJCNN), Anchorage, AL, USA, 14–19 May 2017; pp. 1578–1585. [Google Scholar] [CrossRef]
- Cui, Z.; Chen, W.; Chen, Y. Multi-scale convolutional neural networks for time series classification. arXiv
**2016**, arXiv:1603.06995. [Google Scholar] [CrossRef] - Zheng, Y.; Liu, Q.; Chen, E.; Ge, Y.; Zhao, J.L. Exploiting multi-channels deep convolutional neural networks for multivariate time series classification. Front. Comput. Sci.
**2016**, 10, 96–112. [Google Scholar] [CrossRef] - Tanisaro, P.; Heidemann, G. Time series classification using time warping invariant echo state networks. In Proceedings of the 2016 15th IEEE International Conference on Machine Learning and Applications (ICMLA), Anaheim, CA, USA, 18–20 December 2016; pp. 831–836. [Google Scholar] [CrossRef]
- Karim, F.; Majumdar, S.; Darabi, H.; Chen, S. LSTM fully convolutional networks for time series classification. IEEE Access
**2017**, 6, 1662–1669. [Google Scholar] [CrossRef] - Liu, M.; Ren, S.; Ma, S.; Jiao, J.; Chen, Y.; Wang, Z.; Song, W. Gated transformer networks for multivariate time series classification. arXiv
**2021**, arXiv:2103.14438. [Google Scholar] [CrossRef] - Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A.N.; Kaiser, Ł.; Polosukhin, I. Attention is all you need. Adv. Neural Inf. Process. Syst.
**2017**, 30, 6000–6010. [Google Scholar] [CrossRef] - Zhou, H.; Zhang, S.; Peng, J.; Zhang, S.; Li, J.; Xiong, H.; Zhang, W. Informer: Beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, Washington, DC, USA, 7–14 February 2021; Volume 35, pp. 11106–11115. [Google Scholar] [CrossRef]
- Tsai, Y.H.H.; Bai, S.; Yamada, M.; Morency, L.P.; Salakhutdinov, R. Transformer Dissection: A Unified Understanding of Transformer’s Attention via the Lens of Kernel. arXiv
**2019**, arXiv:1908.11775. [Google Scholar] [CrossRef] - Cao, J.; Chu, J.; Guo, F.; Liu, K.; Xie, R.; Qin, H. Ftmar: A Fusion Transformer Network for Multi-Resident Activity Recognition. SSRN
**2022**. [Google Scholar] [CrossRef] - Baydogan, M.G. Multivariate time series classification datasets. 2015.
- Serrà, J.; Pascual, S.; Karatzoglou, A. Towards a Universal Neural Network Encoder for Time Series. In Proceedings of the International Conference of the Catalan Association for Artificial Intelligence, Alt Empordà, Catalonia, Spain, 8–10 October 2018; pp. 20–129. [Google Scholar] [CrossRef]
- Zhao, B.; Lu, H.; Chen, S.; Liu, J.; Wu, D. Convolutional neural networks for time series classification. J. Syst. Eng. Electron.
**2017**, 28, 162–169. [Google Scholar] [CrossRef] - Le, G.A.; Malinowski, S.; Tavenard, R. Data augmentation for time series classification using convolutional neural networks. In Proceedings of the ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data, Porto, Portugal, 11 September 2016. [Google Scholar]
- Chen, Z.; Zhang, L.; Jiang, C.; Cao, Z.; Cui, W. WiFi CSI based passive human activity recognition using at tention based BLSTM. IEEE Trans. Mob. Comput.
**2018**, 18, 2714–2724. [Google Scholar] [CrossRef]

**Figure 4.**(

**a**) Time Encoder attention map (upper–left); (

**b**) Time Encoder DTW (upper–right); (

**c**) Behavior Encoder attention map (bottom–left); (

**d**) Behavior Encoder L2 distance (bottom–right).

Dataset | Train Cases | Test Cases | Dimensions | Length | Classes |
---|---|---|---|---|---|

AUSLAN | 1140 | 1425 | 22 | 45–136 | 95 |

ArabicDigits | 6600 | 2200 | 13 | 4–93 | 10 |

CMUsubject16 | 29 | 29 | 62 | 127–580 | 2 |

CharacterTrajectories | 300 | 2558 | 3 | 109–205 | 20 |

ECG | 100 | 100 | 2 | 39–152 | 2 |

JapaneseVowels | 270 | 370 | 12 | 7–29 | 9 |

Libras | 180 | 180 | 2 | 45 | 15 |

UWave | 20 | 4278 | 3 | 315 | 8 |

Wafer | 298 | 896 | 6 | 104–198 | 2 |

WalkvsRun | 28 | 16 | 62 | 128–1918 | 2 |

STU | 6159 | 2054 | 13 | 32 | 2 |

**Table 2.**Test accuracy of the TSIN and other benchmark models on the multivariate time-series dataset.

MLP | FCN | ResNet | Encoder | MCNN | t-LeNet | MCDCNN | TCN | TWIESN | GTN | TSIN | |
---|---|---|---|---|---|---|---|---|---|---|---|

AUSLAN | 93.09 | 93.54 | 92.91 | 67.58 | 1.05 | 1.05 | 80.35 | 90.60 | 14.88 | 92.70 | 94.04 |

ArabicDigits | 95.82 | 98.77 | 98.55 | 96.36 | 10.02 | 10.02 | 95.9 | 98.27 | 66.91 | 97.82 | 98.09 |

CMUsubject16 | 96.55 | 93.10 | 93.10 | 89.66 | 53.10 | 51.07 | 55.17 | 62.07 | 65.52 | 100.00 | 100.0 |

CharacterTrajectories | 82.02 | 96.83 | 96.68 | 4.38 | 5.48 | 6.79 | 92.22 | 96.48 | 91.79 | 96.05 | 97.03 |

ECG | 78.00 | 86.00 | 86.00 | 67.00 | 67.00 | 67.00 | 77.00 | 87.00 | 77.00 | 85.00 | 87.00 |

JapaneseVowels | 81.62 | 98.11 | 97.84 | 93.24 | 9.24 | 23.79 | 94.59 | 96.22 | 78.11 | 98.38 | 97.84 |

Libras | 32.22 | 85.00 | 86.67 | 6.67 | 6.67 | 6.67 | 56.11 | 72.78 | 52.22 | 81.67 | 87.22 |

UWave | 87.31 | 91.84 | 89.69 | 12.48 | 12.48 | 12.48 | 86.21 | 90.81 | 51.29 | 88.69 | 88.66 |

Wafer | 90.63 | 97.88 | 95.76 | 96.32 | 89.42 | 89.42 | 65.83 | 98.33 | 87.52 | 97.54 | 98.88 |

WalkvsRun | 70.00 | 100.00 | 100.00 | 100.00 | 75.00 | 60.00 | 45.00 | 100.00 | 87.50 | 100.00 | 100.0 |

STU | 78.09 | 71.23 | 71.86 | 74.68 | 80.77 | 80.32 | 81.30 | 80.22 | 80.72 | 77.90 | 81.45 |

**Table 3.**Ablation study of the Time Encoder, Behavior Encoder, Intermediate Fusion Module and gating in the TSIN.

Time Encoder | Behavior Encoder | Time Encoder + Behavior Encoder + Fusion | Step Encoder + Behavior Encoder + Gate | TSIN | |
---|---|---|---|---|---|

AUSLAN | 89.05 | 91.51 | 87.09 | 91.51 | 94.04 |

ArabicDigits | 96.59 | 98.09 | 97.82 | 98.36 | 98.09 |

CMUsubject16 | 96.55 | 89.66 | 93.10 | 89.66 | 100.00 |

CharacterTrajectories | 91.59 | 96.25 | 95.86 | 96.36 | 97.03 |

ECG | 83.00 | 86.00 | 85.00 | 85.00 | 87.00 |

JapaneseVowels | 94.32 | 96.76 | 96.49 | 96.22 | 97.84 |

Libras | 57.22 | 86.67 | 77.78 | 86.67 | 87.22 |

UWave | 75.57 | 89.29 | 85.55 | 87.89 | 88.66 |

Wafer | 91.74 | 97.99 | 98.44 | 97.66 | 98.88 |

WalkvsRun | 100.00 | 100.00 | 100.00 | 100.00 | 100.00 |

STU | 81.06 | 80.43 | 79.45 | 78.89 | 81.45 |

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |

© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Xu, J.; Ding, X.; Ke, H.; Xu, C.; Zhang, H.
Student Behavior Prediction of Mental Health Based on Two-Stream Informer Network. *Appl. Sci.* **2023**, *13*, 2371.
https://doi.org/10.3390/app13042371

**AMA Style**

Xu J, Ding X, Ke H, Xu C, Zhang H.
Student Behavior Prediction of Mental Health Based on Two-Stream Informer Network. *Applied Sciences*. 2023; 13(4):2371.
https://doi.org/10.3390/app13042371

**Chicago/Turabian Style**

Xu, Jieming, Xuefeng Ding, Hanyu Ke, Cong Xu, and Hanlun Zhang.
2023. "Student Behavior Prediction of Mental Health Based on Two-Stream Informer Network" *Applied Sciences* 13, no. 4: 2371.
https://doi.org/10.3390/app13042371