Smart Neurofeedback System For Meditation

People in the modern society are living a more and more anxious life because of the increasing pressure and the rapid pace of life.Mental illnesses and psychological problems are becoming more common.The mental cultivation methods by LEYI is a collection of scientific and traditional Zen training methods suitable for modern people,which could make people less anxious and feel more blessed.Meditation is an important content for the mental cultivation methods by LEY I,and it is helpful for people to get reach of enlightenment.The scientific discovery of the specific EEG(electroencephalography)pattern of people in meditation could be exploited in the neurofeedback system,which would support people to meditate.In this thesis,EEG got from people in meditation is analyzed using data mining techniques,and a novel calibration method is proposed.The accuracy of classification for novices of meditation get over 90%after the calibration method used on the EEG feature sample.In this thesis,basic concepts and technologies are introduced,neurofeedback therapy and the-state-of-art in meditation researches.From the summary of the researches and products,the drawbacks of them are pointed out.An EEG experiment is designed according to the 4-step meditation method of the mental cultivation methods by LEYI.In the experiment,20 subjects meditate by the sequence of Sit Quietly-Focused Attention-Mindfulness-Transcendent Meditation.The EEG data of Sit Quietly step is used as baseline data to calibrate the data of Focused Attention,Mindfulness and Transcendent Meditation steps.As for data mining work,data segmentation,filtering and data selection are carried out in preprocess phase.EEG data is transformed into 15-dimension feature vectors.Calibration and classifications are conducted.The result shows the calibration method proposed in this thesis could improve the accuracy of the classification algorithms such as random forest,support vector machine and Bayes net,and make the accuracy come up to 90%.To make the neurofeedback application more practical,dimension reduction is tested but the results are unsatisfactory.Based on the EEG experiment and data mining research,an intelligent meditation neurofeedback system is designed and developed.The system is under the idea of supervised learning in which the master sets labels for the trainers and the system give assessments to the trainers based on labeled data.The thesis points out the drawbacks of existed researches and products,and designs the system functions,interfaces and interactions on the need of the mental cultivation methods by LEYI.At the end of the thesis,the problems of enlightenment and the judgement of enlightenment by AI are discussed.Advanced Artificial Intelligence Principle including Full Information theory and the theory of Intelligence Formation is introduced,as well as the theory of Complexity proposed by Stephen Wolfram come from tremendous researches of cellular automata.The thesis analyzes these two theories on the topic of the evaluation of enlightenment by AI.