篇名 | 感應爐溫度分類 |
---|---|
卷期 | 34 |
並列篇名 | Classification of Induction Furnace Temperature |
作者 | 余元利 |
頁次 | 045-054 |
關鍵字 | 多層感知器 、 一位有效編碼 、 準確率 、 Multilayer Perceptron 、 One-hot encoding 、 accuracy |
出刊日期 | 202112 |
本文使用具有兩層隱藏層的多層感知器做為深度學習的模型。標籤集有8個類別,共有230個檔案,總計1745筆,而測試集有36個檔案,總計276筆。資料預處理使每筆的特徵欄位數量為50、標籤欄位數量為8,標籤欄位採用一位有效編碼。標籤集的80%作為訓練、20%作為驗證,評估模型的指標是準確率,分類結果顯示:訓練集與驗證集的準確率分別為100%與99.7%。
This paper uses Multilayer Perceptron with two hidden layers as a deep learning model. The labeled dataset has 8 categories, a total of 230 files, a total of 1745 records. The test dataset has 36 files, a total of 276 records. Data preprocessing makes the number of feature columns for each record 50 and the number of label columns 8. The label columns use One-hot encoding. 80% of the labeled dataset is used for training and 20% is used for validation. The metric of evaluating the model is accuracy. The classification results show that the accuracy of training dataset and validation dataset are 100% and 99.7% respectively.