隨著科技的發展，對於大腦的研究也越來越豐富，認為大腦的結構是具有學習能力，並想試著學習模仿大腦的結構。如果我們可以將機器加入學習能力，學習生活周遭的事情，是否能夠利用機器取代人類，做一些簡單、重複性高的事，讓人類的生活更加輕鬆。
本論文所使用的網路架構是結合非監督式和監督式的類神經網路。透過影像的輸入，將影像結合並傳到下一層皮質區，最後傳達大腦時，由大腦的經驗做判斷。本論文架構採用自我組織映射，將特徵相似度高的放在一起；再透過適應共振理論，對於整個架構的輸出結果做儲存；最後，利用學習向量量化網路，對適應共振理論所儲存的特徵映射響應圖進行微調，而非使用一般常見的梯度修正。另外，我們透過可視覺化，將每層的特徵轉換成具有解讀性特徵影像，使人可以有效的解讀特徵所賦予的意義。
本論文的實驗中，共使用兩種資料集，分別為手寫辨識資料集和歌曲資料集，並對此架構中的特徵映射圖、適應共振理論的警戒參數、影像二值化、影像結合進行比較和結果分析，在最後對特徵圖做可視覺化的呈現。

In recent years, many people try to understand the structure of the brain. We believed that if we can add machines to learning. It is possible to use machines instead of humans and do some simple, repetitive things.
This neural network architecture is a combination of supervised and unsupervised neural networks. This paper uses the k-means algorithm to group the input image. And then use the self-organizing map to generate feature map. And then use adaptive resonance theory to save the results of response map. Finally, use the learning vector quantization network fine-tuning the results of adaptive resonance theory without the use of Gradient. In addition, we can visualize and transform each layer of features into images that can be understood by the human’s eyes.
In this paper’s experiments, we use two types of datasets. One is MNIST, another is song datasets. The experiment includes the feature maps, the alert parameters of the adaptive resonance theory, binary of image, and image combination. In the end, the feature is visualized presented
The experiment in this thesis has compared to the feature maps, the alert parameters of ART, the image binary, and the different method of the image’s merger.

摘 要 i
ABSTRACT ii
誌 謝 iii
目 錄 iv
圖目錄 vii
表目錄 viii
第一章、 緒論 1
1-1 研究動機 1
1-2 研究目的 2
1-3 論文架構 3
第二章、 相關研究 4
2-1 人腦結構 4
2-1-1 人類的大腦 4
2-1-2 記憶 5
2-1-3 皮質的運作 6
2-2 分群演算法 8
2-2-1 相關距離計算方法 8
2-2-2 k-means 演算法 10
2-2-3 側影係數 11
2-2-4 適應共振理論 12
2-3 自我組織映射演算法 14
2-3-1 演算法架構 14
2-4 學習向量量化網路 17
2-4-1 演算法架構 17
2-5 音訊分析 19
2-5-1 頻譜圖 20
2-5-2 梅爾倒頻譜 21
2-5-3 常數Q轉換 22
第三章、 研究方法 23
3-1 相關說明 23
3-1-1 特徵映射圖 23
3-1-2 特徵映射響應圖 24
3-1-3 鄰近函數 27
3-1-4 時間函數 28
3-2 多層自我組織映射 29
3-2-1 架構說明 29
3-2-2 演算法 32
3-3 與卷積神經網路之差異 34
3-3-1 非梯度修正 35
3-3-2 可視覺化之特徵圖 36
第四章、 實驗設計與結果分析 39
4-1 實驗設計 39
4-2 基於多層自我組織映射圖之數字辨識 39
4-2-1 比較不同大小的特徵映射圖 41
4-2-2 比較影像二值化 45
4-2-3 比較不同警戒參數之適應共振理論 45
4-2-4 比較得勝次數的組合方式 46
4-2-5 可視覺化之特徵圖 47
4-2-6 卷積神經網路之比較 53
4-3 基於多層自我組織映射圖之歌曲辨識 55
4-3-1 比較頻譜圖之間的差異 58
4-3-2 比較不同大小的特徵映射圖 60
4-3-3 比較影像二值化 66
4-3-4 比較不同警戒參數之適應共振理論 67
4-3-5 可視覺化之特徵圖 68
4-3-6 卷積神經網路之比較 69
第五章、 結論與展望 71
5-1 結論 71
5-2 未來展望 72
參考資料 73

[1] 洪蘭，創智慧：理解人腦運作，打造智慧機器，初版 編者，臺北市：遠流出版事業股份有限公司，民國九十五年。
[2] V. B. Mountcastle, "An Organizing Principle for Cerebral Function: The Unit Model and the Distributed System," The Mindful Brain, pp. 7-50, 1978.
[3] 蘇木春、張孝德，機器學習：類神經網路、模糊系統以及基因演算法則， 臺北市：全華圖書股份有限公司，民國一百零一年。
[4] P. J. Rousseeuw, "Silhouettes: A graphical aid to the interpretation and validation of cluster analysis," Journal of Computational and Applied Mathematics, vol. 20, pp. 53-65, 1987.
[5] G. A. Carpenter and S. Grossberg, "A massively parallel architecture for a self-organizing neural pattern recognition machine," Computer Vision, Graphics, and Image Processing, vol. 37, pp. 54-115, 1987.
[6] G. A. Carpenter and S. Grossberg, "The ART of adaptive pattern recognition by a self-organizing neural network," Computer, vol. 21, no. 3, pp. 77-88, 1988.
[7] G. A. Carpenter and S. Grossberg, "ART 2: Self-organization of stable category recognition codes for analog input patterns," Applied Optics, vol. 26, pp. 4919-4930, 1987.
[8] M. C. Su and H. T. Chang, "Fast Self-Organizing Feature Map Algorithm," IEEE Transactions on Neural Networks, vol. 11, no. 3, pp. 721-733, 2000.
[9] T. Kohonen, "Improved versions of learning vector quantization," IJCNN, vol. 1, pp. 545-550, 1990.
[10] "Sound," Wikipedia, [Online]. Available: https://en.wikipedia.org/wiki/Sound. [Accessed 17 - May - 2018].
[11] J. Allen, "Short term spectral analysis, synthesis, and modification by discrete Fourier transform," IEEE Transactions on Acoustics, Speech, and Signal Processing, vol. 25, pp. 235-238, 1977.
[12] "Mel-frequency cepstrum," Wikipedia, [Online]. Available: https://en.wikipedia.org/wiki/Mel-frequency_cepstrum. [Accessed 17 - May - 2018].
[13] J. C. Brown, "Calculation of a constant Q spectral transform," The Journal of the Acoustical Society of America, vol. 89, pp. 425-434, 1991.
[14] "Equal temperament," Wikipedia, [Online]. Available: https://en.wikipedia.org/wiki/Equal_temperament. [Accessed 17 - May - 2018].
[15] A. Krizhevsky, I. Sutskever, and G. E. Hinton, "ImageNet Classification with Deep Convolutional Neural Networks," Advances in Neural Information Processing Systems, vol. 1, pp. 1097-1105, 2012.
[16] "Convolutional neural network," Widipedia, [Online]. Available: https://en.wikipedia.org/wiki/Convolutional_neural_network. [Accessed 17 - Jun - 2018].
[17] "神經網絡之父 Geoff Hinton 推翻畢生心血「反向傳播演算法」：打掉重來，AI 才有未來！," [Online]. Available: https://buzzorange.com/techorange/2017/09/22/geoffrey-hinton-fight-back-propagation/. [Accessed 23 - Jul - 2018].
[18] Y. LeCun, C. Cortes, and C. J. Burges, "THE MNIST DATABASE of Handwritten Digits," [Online]. Available: http://yann.lecun.com/exdb/mnist/. [Accessed 7 - Jun - 2018].
[19] "音符時值," Wikiwand, [Online]. Available: http://www.wikiwand.com/zh-hk/音符時值. [Accessed 13 - Jun - 2018].
[20] "藝術與人文本位課程," [Online]. Available: http://www.sses.tc.edu.tw/~journal/journal017/passport/passport1.htm. [Accessed 11 - Jun - 2018].
[21] "梅爾頻率倒譜係數," Wikipedia, [Online]. Available: https://zh.wikipedia.org/wiki/梅爾頻率倒譜係數. [Accessed 13 - Jun - 2018].
[22] "本論文之相關實驗結果," [Online]. Available: https://drive.google.com/drive/folders/1vBebwqeQqoLaGMlJLk2zgVsxR2N7xyo2?usp=sharing. [Accessed 20 - Jun - 2018].
[23] "Vanishing Gradient Problem," Wikipedia, [Online]. Available: https://en.wikipedia.org/wiki/Vanishing_gradient_problem. [Accessed 18 - Jun - 2018].