利用類神經網路與決策樹分析腦波專注值特徵進行認知負荷分類

近年來，因物聯網技術的發展使過去的大型量測腦電圖設備逐漸被開發為穿戴式腦電圖設備，使越來越多教育相關研究者使用該設備量測腦波與專注力。然而在使用該儀器與認知負荷的研究中，至今蒐集認知負荷的方式大多還是使用問卷方式進行蒐集，導致無法完整蒐集到長時間多事件的認知負荷情形。因此，本研究使用穿戴式腦電圖設備蒐集受測者在事件活動時的專注力，並在結束時給予受測者填寫認知負荷問卷。在資料處理時，透過區段切割的方式取出專注力的特徵值進行模型訓練，並利用該模型進行分類之正確率探討。從本研究的結果得知，訓練出的模型在整體正確率時已有良好表現，並在探討各別負荷正確率上低負荷與中負荷也有良好的正確率表現。因此本研究認為使用該方式訓練的模型可以幫助解決執行連續事件時無法仔細填寫認知負荷問卷的問題。

關鍵字

穿戴式腦電圖設備；認知負荷；分類

並列摘要

In recent years, due to the development of the Internet of Things technology, large-scale EEG devices in the past have been gradually developed as wearable EEG devices, which has more and more education-related researchers to use this device to EEG and attention values. However, in researches using this instrument and cognitive load, most of the methods used to collect the cognitive load so far have been collected using questionnaires, resulting in the inability to collect the cognitive load of multiple events for a long time. Therefore, in this study, wearable EEG was used to collect the subject's attention values during event activities, and at the end of the study, subjects were given a cognitive load questionnaire. During data processing, the feature of attention values were extracted by segment cutting for model training, and the correctness of classification were discussed using the model. It is known from the results of this study that the trained model has a good accuracy rate when it does Indiscriminate cognitive loads, and it also has a good accuracy rate performance when discussing the load accuracy rate of low cognitive load and medium cognitive load. Therefore, this study believes that the model trained in this way can help solve the problem of failing to fill in the cognitive load questionnaire carefully when performing continuous events.

並列關鍵字

Wearable EEG Device ； Cognitive Load ； Classification

國際替代計量

利用類神經網路與決策樹分析腦波專注值特徵進行認知負荷分類

全文下載

主題瀏覽