查詢結果分析
來源資料
相關文獻
- The Frequency Sensitive Competitive Learning Algorithms for Vector Quantization
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- Neural Network Procedures for Taguchi's Dynamic Problems
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
- 臺灣汽保費率之估計--對數線性費率模式與類神經網路之比較
- 運用類神經網路於股價指數之套利--以日經225指數為例
- 使用類神經網路預估碳化鎢材料放電加工性能
頁籤選單縮合
題 名 | The Frequency Sensitive Competitive Learning Algorithms for Vector Quantization |
---|---|
作 者 | 呂蘋; 張鴻義; | 書刊名 | 南開學報 |
卷 期 | 2 1997.06[民86.06] |
頁 次 | 頁59+61-71 |
分類號 | 312.2 |
關鍵詞 | 類神經網路; 競爭學習法; 資料壓縮編碼; Vector Quantization; Competitive Learning Algorithm; |
語 文 | 英文(English) |
中文摘要 | 以類神經網路為基礎的競爭學習法可適用於向量式的壓縮編碼,文獻[1}中比較 了數種不同的競爭性學習法在資料壓縮編碼上的效能表現, 同時在文獻 [1] 中提出一種稱 為意識性競爭學習法,這種學習法可讓向量編碼法中的所有字碼具有相同的機會成為競爭的 贏家。亦即它們有相同的機率在編碼的過程中被採用。這種意識性競爭學習法被用於數位語 音壓縮時具有良好的效果。不幸的,當我們將之用於數位影像編碼時,效果卻無傳統的競爭 學習法來得好。本文即以此觀點切入,來討論其可能之原因。 |
英文摘要 | A number of training algorithms for competitive learning networks appl ied to problem of vector quantization for data compression are compared in [1]. They introduced a new competitive learning algorithm based on the 'conscience' learning method. The 'conscience' learning method is designed to achieve that the neurons have the same probability to wine. This method had shown been better than the traditional competitive learning method for speech data compression [1]. Recently, it was mentioned again in [2] to realize a adaptive vector quan tization for image sequence coding. The scheme in [2] to realize a adaptive vector quantization for image sequence coding. The scheme in [2] to realize base d on the codeword refreshing when image data transmitting. Although the applicat ion with 'conscience' learning method for competitive learning in [1] is designed to peroform a speech data compression, we implement it for the image data compression. unfortunately, the MSE of this new learning method is not better than traditional competitive learning for image data compression, although it is better for speech data compression in [1]. We have a discussion about this different between [1] and us. |
本系統中英文摘要資訊取自各篇刊載內容。