使用雲端機器學習預測運動時專注力與放鬆度影響之研究

A Study of the Prediction of Attention and Meditation Using Cloud Machine Learning on Exercise

本文以政府資料開放平台的開放資料（Open Data），探討國人近年來運動習慣狀況，分析運動時的生理狀況。本研究使用腦波儀與行動裝置，搭配自行開發的行動裝置應用程式，擷取與儲存腦波專注力和放鬆度的資料，以預測運動的專注力與放鬆度。希望將科學儀器帶入運動休閒中，達到智慧運動科技的理念。本文使用雲端機器學習平台，將資料透過不同的迴歸演算法進行預測分析。綜合評估演算法模型的決定係數，並給予人們適當的運動方向與建議。實驗結果顯示，在預測專注度的迴歸模型中，以線性迴歸的決定係數0.852817為最高。在預測放鬆度的迴歸模型中，以神經網路迴歸的決定係數0.752456為最高。

This article takes the open data of government information open platform; it explores the exercise of people in recent years, it analyzes the physiological state of exercise. This study uses brainwave and mobile devices to capture and store brainwave attention and meditation information with a self-developed mobile application to predict exercise attention and meditation. Hope that the scientific instruments into the sports and leisure, to the wisdom of the concept of sports technology. This paper uses the cloud machine learning platform to predict and analyze the algorithm by classification and regression, and comprehensively evaluate the decision coefficient of the algorithm model and give the appropriate direction and suggestion of the exercise. Finally, the experimental result shows the machine learning and predicting regression model, the linear regression coefficient of determination is 0.852817.