集成方法在影像辨識中之實驗與討論;Experiment and Discussion of Ensemble MethodOn the Image Recognition

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/81514

Title:	集成方法在影像辨識中之實驗與討論;Experiment and Discussion of Ensemble MethodOn the Image Recognition
Authors:	劉致嘉;Liu, Chih-Chia
Contributors:	數學系
Keywords:	集成學習;支持向量機;卷積神經網路;循環神經網路;unsemble learning;SVM;CNN;RNN
Date:	2019-08-15
Issue Date:	2019-09-03 15:58:17 (UTC+8)
Publisher:	國立中央大學
Abstract:	本文旨在探討機器學習中將不同類型的演算法做組合之後，在影像辨識上對於分類準確度的影響，以及試圖建立一個使用集成方法來分類資料的流程。過程中分別測試了三種類型演算法在資料集上獨立運作的準確度，接著再根據集成學習中的組合方法，對於同類型及不同類型演算法的組合模型進行實驗，最後觀察其結果並進行討論及總結。;The purpose of this thesis is to observe the classification accuracy of image recognition after combine different types of algorithms, and try to build a process about using ensemble method to classify data.In the process, the accuracy of three types of algorithms were tested, then according to the combination method of ensemble, experiment was design by the model which combined with the same or different types of algorithms. Finally, we observe the result and make discuss and summarize.
Appears in Collections:	[Graduate Institute of Mathematics] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	261	View/Open

社群 sharing