查詢結果分析
來源資料
頁籤選單縮合
| 題 名 | A Study of Disaster Prevention Industry Data Analysis=防災產業數據分析之研究 |
|---|---|
| 作 者 | 林建谷; 陳祐祥; 張介仁; 黃鋕展; 林玉華; | 書刊名 | 修平學報 |
| 卷 期 | 50 2025.09[民114.09] |
| 頁 次 | 頁1-17 |
| 分類號 | 312.83 |
| 關鍵詞 | 大數據分析; 機器學習; 人工智慧; Big data analysis; Machine learning; Artificial intelligence; |
| 語 文 | 英文(English) |
| 中文摘要 | 智慧製造的關鍵是透過物聯網與感測相關技術從大數據資料進行分析,改善機械與人類之間相互溝通的效率。本研究使用防災產業的產品數據進行分析,研究如何預測產品的相關參數數據等,以提高生產績效。研究方法使用機器學習的3種分類器模型以預測產品的品質。研究結果顯示:(1)各個分類器的預測準確率最高數據依序為Random forests(90.9%)、K-nearest neighbors(90.8%)、Neural networks(90.7%)。(2)可以預測出15.5%~19.9%的相關產品參數輸入有問題。研究結果可以提供和協助相關企業改善生產績效之研究參考。 |
| 英文摘要 | In order to increase the effectiveness of communication between machines and humans, big data analysis via the Internet of Things and sensing-related technologies is essential to smart manufacturing. This study uses product data from disaster prevention industry for analysis and studies how to predict related product parameter data in order to improve production performance. The research method uses three classification models of machine learning to predict the related product quality. Research results indicate that (1) the classifiers with the highest prediction accuracy are random forests (90.9%), K-nearest neighbors (90.8%), and neural networks (90.7%). (2) It can be predicted that 15.5%~19.9% of product parameter inputs have problems in advance. The research results can help companies improve their production performance as a research reference. |
本系統中英文摘要資訊取自各篇刊載內容。