EEG Sinyalleri Kullanılarak Makine Öğrenmesi Tabanlı Otomatik Duygu Sınıflandırma

Year 2023, Volume: 1 Issue: 1, 26 - 40, 30.06.2023

Hakan Köksal , Mehmet Bayğın

https://doi.org/10.5281/zenodo.8074861

Abstract

Son yıllarda otomatik duygu tespiti ve sınıflandırma literatürde üzerinde çalışılan konular arasında yer almaktadır. Duygular, bireylerin dış dünyayla olan ilişkilerinde, eylemlerinde ve kararlarında etkin rol oynamaktadır. Bu nedenle duygu tanıma insan-bilgisayar etkileşimde kritik öneme sahiptir. Duyguların tespiti yapılırken EEG sinyallerinin bazı nörolojik ve beyinsel aktiviteleri tespit ettiği tartışma konusudur. Bu çalışmada, duyguların tespiti ve analizi için ayırıcı özellik taşıyan sinyaller üretildiğinden EEG sinyalleri kullanılmıştır. Bu çalışmada etkili ve basit yöntemler geliştirerek yüksek doğruluğa sahip otomatik duygu tanıma amaçlanmaktadır. Önerilen Local Binary Pattern (LBP) yönteminde ReliefF özellik seçimi ve Ensemble sınıflandırıcı kullanılarak yeni bir otomatik EEG duygu tanıma modeli sunulmuştur. Bu model, özellik çıkarma, özellik seçme ve sınıflandırma olmak üzere makine öğrenimi modelinin tüm aşamalarını kapsamaktadır. Bu model üzerinde yapılan çalışmada, DREAMER veri seti üzerinden ortalama %63.89 sınıflandırma başarısına ulaşılmıştır.

Keywords

EEG, Makine Öğrenmesi, Duygu Tanıma, Sınıflandırma

Machine Learning Based Automatic Emotion Classification Using EEG Signals

Abstract

In recent years, automatic emotion detection and classification are among the topics studied in the literature. Emotions play an active role in individuals' relations with the outside world, their actions and decisions. Therefore, emotion recognition is critical in human-computer interaction. It is a matter of debate that EEG signals detect some neurological and cerebral activities while detecting emotions. In this study, EEG signals were used as distinctive signals were produced for the detection and analysis of emotions. In this study, automatic emotion recognition with high accuracy is aimed by developing effective and simple methods. In the proposed Local Binary Pattern (LBP) method, a new automatic EEG emotion recognition model is presented using ReliefF feature selection and Ensemble classifier. This model covers all phases of the machine learning model, including feature extraction, feature selection, and classification. In the study conducted on this model, an average of 63.89% classification success was achieved over the DREAMER data set.

Keywords

EEG, Machine Learning, Emotion Recognition, Classification

References

• B. Parkinson, “Emotions are social”, Br. J. Psychol., c. 87, sayı 4, ss. 663–683, 1996, doi: 10.1111/j.2044-8295.1996.tb02615.x.
• S. Liu, Z. Wang, Y. An, J. Zhao, Y. Zhao, ve Y. D. Zhang, “EEG emotion recognition based on the attention mechanism and pre-trained convolution capsule network”, Knowledge-Based Syst., c. 265, s. 110372, 2023, doi: 10.1016/j.knosys.2023.110372.
• J. Prinz, “Which emotions are basic?”, Emot. Evol. Ration., ss. 1–19, 2012, doi: 10.1093/acprof:oso/9780198528975.003.0004.
• D. Maheshwari, S. K. Ghosh, R. K. Tripathy, M. Sharma, ve U. R. Acharya, “Automated accurate emotion recognition system using rhythm-specific deep convolutional neural network technique with multi-channel EEG signals”, Comput. Biol. Med., c. 134, sayı May, s. 104428, 2021, doi: 10.1016/j.compbiomed.2021.104428.
• R. Jenke, A. Peer, ve M. Buss, “Feature Extraction and Selection for Emotion Recognition from Electrodermal Activity”, IEEE Trans. Affect. Comput., c. 12, sayı 4, ss. 857–869, 2021, doi: 10.1109/TAFFC.2019.2901673.
• A. Dogan vd., “Automated accurate emotion classification using Clefia pattern-based features with EEG signals”, Int. J. Healthc. Manag., ss. 1–14, 2022, doi: 10.1080/20479700.2022.2141694.
• G. Xiao, M. Shi, M. Ye, B. Xu, Z. Chen, ve Q. Ren, “4D attention-based neural network for EEG emotion recognition”, Cogn. Neurodyn., ss. 1–14, 2022, doi: 10.1007/s11571-021-09751-5.
• J. Li vd., “Cross-subject EEG emotion recognition combined with connectivity features and meta-transfer learning”, Comput. Biol. Med., c. 145, sayı April, s. 105519, 2022, doi: 10.1016/j.compbiomed.2022.105519.
• M. yu Zhong, Q. yu Yang, Y. Liu, B. yu Zhen, F. da Zhao, ve B. bei Xie, “EEG emotion recognition based on TQWT-features and hybrid convolutional recurrent neural network”, Biomed. Signal Process. Control, c. 79, sayı P2, s. 104211, 2023, doi: 10.1016/j.bspc.2022.104211.
• A. R. Aguiñaga, L. M. Delgado, V. R. López-López, ve A. C. Téllez, “EEG-Based Emotion Recognition Using Deep Learning and M3GP”, Appl. Sci., c. 12, sayı 5, s. 2527, 2022, doi: doi.org/10.3390/.
• Z. Zhang, S. hua Zhong, ve Y. Liu, “GANSER: A Self-supervised Data Augmentation Framework for EEG-based Emotion Recognition”, IEEE Trans. Affect. Comput., c. XX, sayı XX, ss. 1–17, 2022, doi: 10.1109/TAFFC.2022.3170369.
• R. Yuvara, P. Thagavel, J. Thomas, ve J. Fogarty, “Comprehensive Analysis of Feature Extraction Methods for Emotion Recognition from Multichannel EEG Recordings”, Sensors, ss. 1–19, 2023, doi: 10.3390/s23020915.
• Y. Wei, Y. Liu, C. Li, J. Cheng, R. Song, ve X. Chen, “TC-Net : A Transformer Capsule Network for EEG-based emotion recognition”, Comput. Biol. Med., c. 152, sayı June 2022, s. 106463, 2023, doi: 10.1016/j.compbiomed.2022.106463.
• K. R. Scherer, “What are emotions? and how can they be measured?”, Soc. Sci. Inf., c. 44, sayı 4, ss. 695–729, 2005, doi: 10.1177/0539018405058216.
• L. F. Barrett, M. Gendron, ve Y. M. Huang, “Do discrete emotions exist?”, Philos. Psychol., c. 22, sayı 4, ss. 427–437, 2009, doi: 10.1080/09515080903153634.
• E. Harmon-Jones, C. Harmon-Jones, ve E. Summerell, “On the importance of both dimensional and discrete models of emotion”, Behav. Sci. (Basel)., c. 7, sayı 4, 2017, doi: 10.3390/bs7040066.
• M. M. Bradley ve P. J. Lang, “Measuring emotion: The self-assessment manikin and the semantic differential”, J. Behav. Ther. Exp. Psychiatry, c. 25, sayı 1, ss. 49–59, Mar. 1994, doi: 10.1016/0005-7916(94)90063-9.
• S. Katsigiannis ve N. Ramzan, “DREAMER: A Database for Emotion Recognition Through EEG and ECG Signals from Wireless Low-cost Off-the-Shelf Devices”, IEEE J. Biomed. Heal. Informatics, c. 22, sayı 1, ss. 98–107, 2018, doi: 10.1109/JBHI.2017.2688239.
• I. W. Selesnick, “Wavelet transform with tunable Q-factor”, IEEE Trans. Signal Process., c. 59, sayı 8, ss. 3560–3575, 2011, doi: 10.1109/TSP.2011.2143711.
• M. Baygin, “An accurate automated schizophrenia detection using TQWT and statistical moment based feature extraction”, Biomed. Signal Process. Control, c. 68, sayı January, s. 102777, 2021, doi: 10.1016/j.bspc.2021.102777.
• J. Cheng vd., “Emotion Recognition From Multi-Channel”, IEEE J. Biomed. Heal. Informatics, c. 25, sayı 2, ss. 453–464, 2021.
• Y. Liu vd., “Multi-channel EEG-based emotion recognition via a multi-level features guided capsule network”, Comput. Biol. Med., c. 123, sayı July, s. 103927, 2020, doi: 10.1016/j.compbiomed.2020.103927.
• A. Bhattacharyya, R. K. Tripathy, ve L. Garg, “A Novel Multivariate-Multiscale Approach for Computing EEG Spectral and Temporal Complexity for Human Emotion Recognition”, c. 21, sayı 3, ss. 3579–3591, 2021, doi: 10.1109/JSEN.2020.3027181.
• Y. Wang, S. Qiu, X. Ma, ve H. He, “A prototype-based SPD matrix network for domain adaptation EEG emotion recognition”, Pattern Recognit., c. 110, s. 107626, 2021, doi: 10.1016/j.patcog.2020.107626.
• A. Dogan vd., “PrimePatNet87: Prime pattern and tunable q-factor wavelet transform techniques for automated accurate EEG emotion recognition”, Comput. Biol. Med., c. 138, sayı September, s. 104867, 2021, doi: 10.1016/j.compbiomed.2021.104867.
• T. Tuncer, S. Dogan, ve A. Subasi, “LEDPatNet19: Automated emotion recognition model based on nonlinear LED pattern feature extraction function using EEG signals”, Cogn. Neurodyn., c. 16, sayı 4, ss. 779–790, 2022, doi: 10.1007/s11571-021-09748-0.
• C. Li vd., “Emotion recognition from EEG based on multi-task learning with capsule network and attention mechanism”, Comput. Biol. Med., c. 143, sayı January, s. 105303, 2022, doi: 10.1016/j.compbiomed.2022.105303.
• T. Tuncer, S. Dogan, M. Baygin, ve U. Rajendra Acharya, “Tetromino pattern based accurate EEG emotion classification model”, Artif. Intell. Med., c. 123, sayı March 2021, s. 102210, 2022, doi: 10.1016/j.artmed.2021.102210.
• Z. He, Y. Zhong, ve J. Pan, “Joint Temporal Convolutional Networks and Adversarial Discriminative Domain Adaptation for Eeg-Based Cross-Subject Emotion Recognition”, ICASSP, IEEE Int. Conf. Acoust. Speech Signal Process. - Proc., c. 2022-May, ss. 3214–3218, 2022, doi: 10.1109/ICASSP43922.2022.9746600.
• J. Quan, Y. Li, L. Wang, R. He, S. Yang, ve L. Guo, “EEG-based cross-subject emotion recognition using multi-source domain transfer learning”, Biomed. Signal Process. Control, c. 84, sayı September 2022, s. 104741, 2023, doi: 10.1016/j.bspc.2023.104741.