消费者偏好预测的深度学习神经网络模型
|
摘要
通过对于人类大脑活动的研究来分析消费者对广告和产品的反应的神经营销正在受到新的关注。针对基于脑电波(EEG)的神经营销,提出了一种基于深度学习神经网络的消费者对产品的偏好预测方法。首先,为了提取消费者EEG的特征,采用短时傅里叶变换(STFT)与双调和样条插值,从多通道脑电信号中得到了5个不同频带的EEG形图视频;然后,提出了一种结合5个三维卷积神经网络(3D CNN)与多层长短期记忆(LSTM)神经网络的预测模型,用于从脑电地形图视频预测到消费者的偏好。与卷积神经网络(CNN)模型和LSTM神经网络模型相比,消费者依赖模型的平均准确度分别提高了15.05个百分点和19.44个百分点,消费者独立模型的平均准确度分别提高了16.34个百分点和17.88个百分点。理论分析与实验结果表明,所提出的消费者偏好预测系统可以以低成本提供有效的营销策略开发和营销管理。
Neuromarketing, by which consumer responses to advertisements and products are analyzed through research on human brain activity, is receiving new attention. Aiming at neuromarketing based on ElectroEncephaloGraphy(EEG), a method of consumer preference prediction based on deep learning neural network was proposed. Firstly, in order to extract features of consumer's EEG, five different frequency bands of EEG topographic videos were obtained from multi-channel EEG signals by using Short Time Fourier Transform(STFT) and biharmonic spline interpolation. Then, a prediction model combining five three-Dimensional Convolutional Neural Networks(3 D CNNs) and multi-layer Long Short-Term Memory(LSTM) neural networks was proposed for predicting consumer preference from EEG topographic videos. Compared with the Convolutional Neural Network(CNN) model and LSTM neural network model, the average accuracy of consumer-dependence model was increased by 15.05 percentage points and 19.44 percentage points respectively, and the average accuracy of consumer-independence model was increased by 16.34 percentage points and 17.88 percentage points respectively. Theoretical analysis and experimental results show that the proposed consumer preference prediction system can provide effective marketing strategy development and marketing management at low cost.
Neuromarketing, by which consumer responses to advertisements and products are analyzed through research on human brain activity, is receiving new attention. Aiming at neuromarketing based on ElectroEncephaloGraphy(EEG), a method of consumer preference prediction based on deep learning neural network was proposed. Firstly, in order to extract features of consumer's EEG, five different frequency bands of EEG topographic videos were obtained from multi-channel EEG signals by using Short Time Fourier Transform(STFT) and biharmonic spline interpolation. Then, a prediction model combining five three-Dimensional Convolutional Neural Networks(3 D CNNs) and multi-layer Long Short-Term Memory(LSTM) neural networks was proposed for predicting consumer preference from EEG topographic videos. Compared with the Convolutional Neural Network(CNN) model and LSTM neural network model, the average accuracy of consumer-dependence model was increased by 15.05 percentage points and 19.44 percentage points respectively, and the average accuracy of consumer-independence model was increased by 16.34 percentage points and 17.88 percentage points respectively. Theoretical analysis and experimental results show that the proposed consumer preference prediction system can provide effective marketing strategy development and marketing management at low cost.
引文
[1] SOLEYMANI M,CHANEL G,KIERKELS J J,et al.Affective ranking of movie scenes using physiological signals and content analysis[C]// Proceedings of the 2nd ACM Workshop on Multimedia Semantics.New York:ACM,2008:32-39.
[2] TELPAZ A,WEBB R,LEVY D J.Using EEG to predict consumers' future choices[J].Journal of Marketing Research,2015,52(4):511-529.
[3] AMBLER T,BRAEUTIGAM S,STINS J,et al.Salience and choice:neural correlates of shopping decisions[J].Psychology and Marketing,2004,21(4):247-261.
[4] BALDO D,PARIKH H,PIU Y,et al.Brain waves predict success of new fashion products:a practical application for the footwear retailing industry[J].Journal of Creating Value,2015,1(1):61-71.
[5] KHUSHABA R N,GREENACRE L,KODAGODA S,et al.Choice modeling and the brain:a study on the electroencephalogram (EEG) of preferences[J].Expert Systems with Applications,2012,39(16):12378-12388.
[6] MURUGAPPAN M.Wireless EEG signals based neuromarketing system using fast Fourier transform (FFT)[C]// CSPA 2014:Proceedings of the 10th International Colloquium on Signal Processing and Its Applications.Piscataway,NJ:IEEE,2014:25-30.
[7] BOKSEM M A,SMIDTS A.Brain responses to movie trailers predict individual preferences for movies and their population-wide commercial success[J].Journal of Marketing Research,2015,52(4):482-492.
[8] KAWASAKI M,YAMAGUCHI Y.Effects of subjective preference of colors on attention-related occipital theta oscillations[J].Neuroimage,2012,59(1):808-814.
[9] FAUST O,HAGIWARA Y,HONG T J,et al.Deep learning for healthcare applications based on physiological signals:A review[J].Computer Methods and Programs in Biomedicine,2018,161:1-13.
[10] SHU L,XIE J Y,YANG M Y,et al.A review of emotion recognition using physiological signals[J].Sensors,2018,18(7):2074.
[11] LIANG J W,CHAUDHURI S R,SHINOZUKA M.Simulation of nonstationary stochastic processes by spectral representation[J].Journal of Engineering Mechanics,2007,133(6):616-627.
[12] HUANG G Q,SU Y W,KAREEM A,et al.Time-frequency analysis of nonstationary process based on multivariate empirical mode decomposition[J].Journal of Engineering Mechanics,2016,142(1):04015065.
[13] Mathworks Inc.Griddata function description[EB/OL].[2018- 11- 20].https://ww2.mathworks.cn/help/matlab/ref/griddata.html.
[14] TRAN D,BOURDEV L,FERGUS R,et al.Learning spatiotemporal features with 3D convolutional networks[C]// ICCV 2015:Proceedings of the 2015 IEEE International Conference on Computer Vision.Piscataway,NJ:IEEE,2015:4489-4497.
[15] 孔祥浩,马琳,李海峰,等.CNN与CSP相结合的脑电特征提取与识别方法研究[J].信号处理,2018,34(2):164-173.(KONG X H,MA L,LI H F,et al.Research on EEG feature extraction and recognition method based on CNN and CSP[J].Journal of Signal Processing,2018,34(2):164-173.)
[16] 谢逸,饶文碧,段鹏飞,等.基于CNN和LSTM混合模型的中文词性标注[J].武汉大学学报(理学版),2017,63(3):246-250.(XIE Y,RAO W B,DUAN P F,et al.A Chinese POS tagging approach using CNN and LSTM-based hybrid model[J].Wuhan University Journal of Natural Sciences,2017,63(3):246-250.)
[17] GREFF K,SRIVASTAVA R K,KOUTNíK J,et al.LSTM:a search space odyssey[J].IEEE Transactions on Neural Networks and Learning Systems,2015,28(10):2222-2232.
[18] GERS F A,ECK D,SCHMIDHUBER J.Applying LSTM to time series predictable through time-window approaches[C]// ICANN 2001:Proceedings of the 10th International Conference on Artificial Neural Networks.Berlin:Springer,2001:21-25.
[2] TELPAZ A,WEBB R,LEVY D J.Using EEG to predict consumers' future choices[J].Journal of Marketing Research,2015,52(4):511-529.
[3] AMBLER T,BRAEUTIGAM S,STINS J,et al.Salience and choice:neural correlates of shopping decisions[J].Psychology and Marketing,2004,21(4):247-261.
[4] BALDO D,PARIKH H,PIU Y,et al.Brain waves predict success of new fashion products:a practical application for the footwear retailing industry[J].Journal of Creating Value,2015,1(1):61-71.
[5] KHUSHABA R N,GREENACRE L,KODAGODA S,et al.Choice modeling and the brain:a study on the electroencephalogram (EEG) of preferences[J].Expert Systems with Applications,2012,39(16):12378-12388.
[6] MURUGAPPAN M.Wireless EEG signals based neuromarketing system using fast Fourier transform (FFT)[C]// CSPA 2014:Proceedings of the 10th International Colloquium on Signal Processing and Its Applications.Piscataway,NJ:IEEE,2014:25-30.
[7] BOKSEM M A,SMIDTS A.Brain responses to movie trailers predict individual preferences for movies and their population-wide commercial success[J].Journal of Marketing Research,2015,52(4):482-492.
[8] KAWASAKI M,YAMAGUCHI Y.Effects of subjective preference of colors on attention-related occipital theta oscillations[J].Neuroimage,2012,59(1):808-814.
[9] FAUST O,HAGIWARA Y,HONG T J,et al.Deep learning for healthcare applications based on physiological signals:A review[J].Computer Methods and Programs in Biomedicine,2018,161:1-13.
[10] SHU L,XIE J Y,YANG M Y,et al.A review of emotion recognition using physiological signals[J].Sensors,2018,18(7):2074.
[11] LIANG J W,CHAUDHURI S R,SHINOZUKA M.Simulation of nonstationary stochastic processes by spectral representation[J].Journal of Engineering Mechanics,2007,133(6):616-627.
[12] HUANG G Q,SU Y W,KAREEM A,et al.Time-frequency analysis of nonstationary process based on multivariate empirical mode decomposition[J].Journal of Engineering Mechanics,2016,142(1):04015065.
[13] Mathworks Inc.Griddata function description[EB/OL].[2018- 11- 20].https://ww2.mathworks.cn/help/matlab/ref/griddata.html.
[14] TRAN D,BOURDEV L,FERGUS R,et al.Learning spatiotemporal features with 3D convolutional networks[C]// ICCV 2015:Proceedings of the 2015 IEEE International Conference on Computer Vision.Piscataway,NJ:IEEE,2015:4489-4497.
[15] 孔祥浩,马琳,李海峰,等.CNN与CSP相结合的脑电特征提取与识别方法研究[J].信号处理,2018,34(2):164-173.(KONG X H,MA L,LI H F,et al.Research on EEG feature extraction and recognition method based on CNN and CSP[J].Journal of Signal Processing,2018,34(2):164-173.)
[16] 谢逸,饶文碧,段鹏飞,等.基于CNN和LSTM混合模型的中文词性标注[J].武汉大学学报(理学版),2017,63(3):246-250.(XIE Y,RAO W B,DUAN P F,et al.A Chinese POS tagging approach using CNN and LSTM-based hybrid model[J].Wuhan University Journal of Natural Sciences,2017,63(3):246-250.)
[17] GREFF K,SRIVASTAVA R K,KOUTNíK J,et al.LSTM:a search space odyssey[J].IEEE Transactions on Neural Networks and Learning Systems,2015,28(10):2222-2232.
[18] GERS F A,ECK D,SCHMIDHUBER J.Applying LSTM to time series predictable through time-window approaches[C]// ICANN 2001:Proceedings of the 10th International Conference on Artificial Neural Networks.Berlin:Springer,2001:21-25.