學術產出-Theses
Article View/Open
Publication Export
-
題名 基於古典與量子卷積神經網絡的時間序列圖像分析
Time Series Image Analysis by Classical and Quantum Convolutional Neural Networks作者 周琪
Zhou, Qi貢獻者 廖四郎
Liao, Szu-Lang
周琪
Zhou, Qi關鍵詞 時間序列圖像化
古典卷積神經網路
量子卷積神經網絡
圖像分類
趨勢預測
Imaging time series
Classical convolutional neural network
Quantum convolutional neural network
Image classification
Trend prediction日期 2020 上傳時間 3-Aug-2020 17:39:47 (UTC+8) 摘要 時間序列是一維數據，但我們可以將其轉換成二維矩陣，最終圖像化後成三維張量。圖像化時間序列的映射需要能夠保留時間依賴性等時間序列重要特徵。卷積神經網絡是深度學習中一種重要的神經網絡，在計算機視覺領域有著非常多的應用，其對視覺信息的處理能力非常突出。所以我們將時間序列圖像和卷積神經網絡結合，研究從高維數據上提取數據特徵的可行性與能力，並最終用來進行時間序列圖像分類和未來趨勢預測。研究結果顯示，卷積神經網絡在時間序列圖像分類和未來趨勢預測上有良好表現，其策略表現可以超越基準線並獲得正的累積收益。另外，我們也對量子卷積神經網絡做了初步探討，我們闡述了相關理論，同時模擬構建了一個量子卷積神經網絡模型。研究結果顯示，量子卷積神經網絡是可行有效的，且在時間序列圖像分類和未來趨勢預測上能夠達到古典卷積神經網絡的水準。量子卷積神經網絡具有很大潛力且值得被研究。
Time series is a one dimensional data, but we can transform it to be a two dimensional matrix, and finally obtain a time series image, which is a three dimensional tensor. A map of imaging time series should preserve some important properties of the time series such as the temporal dependency. Convolutional neural network is a kind of neural networks in deep learning, it is widely applied in the field of computer vision for its strong visual information processing ability. Therefore, we combine the time series image and the convolutional neural network together to research the feasibility and the ability of extracting features from the high dimensional data, and use the model to do the time series image classification and the future trend prediction. The result shows that the convolutional neural network has a good performance on time series image classification and future trend prediction. It could beat the baseline and has a positive cumulative return. In addition, we do a preliminary research on the quantum convolutional neural network. We describe the relative theory and simulate a quantum convolutional neural network model. The result shows that the quantum convolutional neural network is feasible and could reach the similar level of the classical convolutional neural network on both time series image classification and future trend prediction. Quantum convolutional neural network is potential and deserves to be studied.參考文獻 [1] Cong, I., Choi, S., & Lukin, M. D. (2019). Quantum Convolutional Neural Networks. Nature Physics, 15, 1273-1278.[2] Eckman, J. P., Kamphorst, S. O., & Ruelle, D. (1987). Recurrence Plots of Dynamical Systems. Europhysics Letters, 4 (91), 973-977.[3] Erhan, D., Bengio, Y., Courville, A., Manzagol, P. A., & Vincent, P. (2010). Why Does Unsupervised Pre-training Help Deep Learning? Journal of Machine Learning Research, 11, 625-660.[4] Heaton, J. B., Polson, N. G., & Witte, J. H. (2016). Deep Learning for Finance: Deep Portfolios. Applied Stochastic Models in Business and Industry, 33, 3-12.[5] Heaton, J. B., Polson, N. G., & Witte, J. H. (2016). Deep Portfolio Theory. arXiv:1605.07230.[6] Kavukcuoglu, K., Sermanet, P., Boureau, Y. L., Gregor, K., Mathieu, M., & LeCun, Y. (2010). Learning Convolutional Feature Hierarchies for Visual Recognition. Neural Information Processing Systems, 1, 1090-1098.[7] Kaye, P., Laflamme, R., & Mosca, M. (2019). An Introduction to Quantum Computing. Oxford University Press.[8] Kerenidis, I., & Prakash, A. (2016). Quantum Recommendation Systems. Innovations in Theoretical Computer Science Conference, 49, 1-21.[9] Kerenidis, I., Landman, J., Luongo, A., & Prakash, A. (2018). Q-means: A Quantum Algorithm for Unsupervised Machine Learning. Neural Information Processing Systems.[10] Kerenidis, I., Landman, J., & Prakash, A. (2020). Quantum Algorithms for Deep Con- volutional Neural Network. International Conference on Learning Representations.[11] Kitaev, A. Y., Shen, A. H., & Vyalyi, M. N. (1999). Classical and Quantum Computa- tion. American Mathematical Society.[12] Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017). ImageNet Classification with Deep Convolutional Neural Networks. Communications of the Association for Computing Machinery, 60(6), 84-90.[13] Lai, C. (2018). Analysis of the predictive ability of time series using convolutional neural network. National Cheng-Chi University.[14] Le, Q. V., Ngiam, J., Chen, Z., Chia, D., Koh, P. W., & Ng, A. Y. (2010). Tiled Convolutional Neural Networks. Neural Information Processing Systems, 1, 1279-1287.[15] LeCun, Y., & Bengio, Y. (1995). Convolutional Networks for Images, Speech, and Time Series. The Handbook of Brain Theory and Neural Networks, 255-258.[16] LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based Learning Ap- plied to Document Recognition. Proceedings of the Institute of Electrical and Electronics Engineers, 86(11), 2278-2324.[17] Martin, T., Hagan, M. T., Demuth, H. B., Beale, M. H., & Jesús, O. D. (2014). Neural Network Design. Martin Hagan.[18] Nakahara, M., & Ohmi, T. (2008). Quantum Ccomputing From Linear Algebra to Physical Realizations. Chemical Rubber Company Press.[19] Nielsen, M. A., & Chuang, I. L. (2010). Quantum Computation and Quantum Infor- mation. University Press of Cambridge.[20] Sakurai, J. J., & Napolitano, J.J. (2014). Modern Quantum Mechanics. Pearson Edu- cation Limited.[21] Scherer, W. (2019). Mathematics of Quantum Computing. Springer Nature Switzerland AG.[22] Shankar, R. (1994). Principles of Quantum Mechanics. Plenum Press.[23] Susskind, L., & Friedman, A. (2014). Quantum Mechanics, The Theoretical Minimum. Perseus Books Group.[24] Wang, Z., & Oates, T. (2015). Imaging Time-Series to Improve Classification and Impu- tation. Proceedings of the International Conference on Artificial Intelligence, 3939-3945.[25] Wu, J. (2017). Introduction to Convolutional Neural Networks. Nanjing University.[26] Zettili, N. (2009). Quantum Mechanics Concepts and Applications. John Wiley & Sons Limited. 描述 碩士
國立政治大學
金融學系
107352038資料來源 http://thesis.lib.nccu.edu.tw/record/#G0107352038 資料類型 thesis dc.contributor.advisor 廖四郎 zh_TW dc.contributor.advisor Liao, Szu-Lang en_US dc.contributor.author (Authors) 周琪 zh_TW dc.contributor.author (Authors) Zhou, Qi en_US dc.creator (作者) 周琪 zh_TW dc.creator (作者) Zhou, Qi en_US dc.date (日期) 2020 en_US dc.date.accessioned 3-Aug-2020 17:39:47 (UTC+8) - dc.date.available 3-Aug-2020 17:39:47 (UTC+8) - dc.date.issued (上傳時間) 3-Aug-2020 17:39:47 (UTC+8) - dc.identifier (Other Identifiers) G0107352038 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/130998 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 金融學系 zh_TW dc.description (描述) 107352038 zh_TW dc.description.abstract (摘要) 時間序列是一維數據，但我們可以將其轉換成二維矩陣，最終圖像化後成三維張量。圖像化時間序列的映射需要能夠保留時間依賴性等時間序列重要特徵。卷積神經網絡是深度學習中一種重要的神經網絡，在計算機視覺領域有著非常多的應用，其對視覺信息的處理能力非常突出。所以我們將時間序列圖像和卷積神經網絡結合，研究從高維數據上提取數據特徵的可行性與能力，並最終用來進行時間序列圖像分類和未來趨勢預測。研究結果顯示，卷積神經網絡在時間序列圖像分類和未來趨勢預測上有良好表現，其策略表現可以超越基準線並獲得正的累積收益。另外，我們也對量子卷積神經網絡做了初步探討，我們闡述了相關理論，同時模擬構建了一個量子卷積神經網絡模型。研究結果顯示，量子卷積神經網絡是可行有效的，且在時間序列圖像分類和未來趨勢預測上能夠達到古典卷積神經網絡的水準。量子卷積神經網絡具有很大潛力且值得被研究。 zh_TW dc.description.abstract (摘要) Time series is a one dimensional data, but we can transform it to be a two dimensional matrix, and finally obtain a time series image, which is a three dimensional tensor. A map of imaging time series should preserve some important properties of the time series such as the temporal dependency. Convolutional neural network is a kind of neural networks in deep learning, it is widely applied in the field of computer vision for its strong visual information processing ability. Therefore, we combine the time series image and the convolutional neural network together to research the feasibility and the ability of extracting features from the high dimensional data, and use the model to do the time series image classification and the future trend prediction. The result shows that the convolutional neural network has a good performance on time series image classification and future trend prediction. It could beat the baseline and has a positive cumulative return. In addition, we do a preliminary research on the quantum convolutional neural network. We describe the relative theory and simulate a quantum convolutional neural network model. The result shows that the quantum convolutional neural network is feasible and could reach the similar level of the classical convolutional neural network on both time series image classification and future trend prediction. Quantum convolutional neural network is potential and deserves to be studied. en_US dc.description.tableofcontents 1 Introduction 11.1 Background and motivation 11.2 Purpose 21.3 Article structure 22 Literature review 32.1 Imaging time series 32.2 Neural network 42.3 Quantum theory 63 Imaging time series 93.1 Tensor representation 93.2 Gramian angular field (GAF) 103.3 Markov transition field (MTF) 133.4 Recurrence plot (RP) 153.5 Merged image 164 Data processing 184.1 Time series 184.2 Time series image 194.3 Data set 235 Classical convolutional neural network (CCNN) 245.1 Forward pass 245.1.1 Convolutional layer 245.1.2 Activation function 255.1.3 Pooling layer 265.1.4 Fully connected layer 275.2 Backward pass 275.2.1 Loss function 275.2.2 Optimization 286 Quantum convolutional neural network (QCNN) 306.1 Quantum information and quantum computing 306.2 Forward pass 326.2.1 Convolution as matrix multiplication 326.2.2 Quantum convolution 356.2.3 Activation function 366.2.4 Quantum sampling 386.2.5 Pooling layer 406.2.6 Fully connected layer 406.3 Backward pass 406.3.1 Loss function 406.3.2 Optimization 417 Experimental design 427.1 Convolutional neural network architecture 427.2 Strategy 447.3 Evaluation 447.3.1 Confusion matrix 447.3.2 Annual return rate 457.3.3 Max drawdown 457.3.4 Calmar ratio 457.4 Experimental groups 468 Research result 478.1 Result of the train and validation sets 478.2 Result of the test set 508.3 Performance of the strategy 539 Conclusion, discussion, and future work 659.1 Conclusion 659.2 Discussion 669.2.1 About the merged data 669.2.2 About the label of image 669.2.3 About the magnitude of the data set 679.2.4 About the normalization and activation function 689.2.5 About the predictions of convolutional neural networks 689.3 Future work 69References 70 zh_TW dc.format.extent 3074316 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0107352038 en_US dc.subject (關鍵詞) 時間序列圖像化 zh_TW dc.subject (關鍵詞) 古典卷積神經網路 zh_TW dc.subject (關鍵詞) 量子卷積神經網絡 zh_TW dc.subject (關鍵詞) 圖像分類 zh_TW dc.subject (關鍵詞) 趨勢預測 zh_TW dc.subject (關鍵詞) Imaging time series en_US dc.subject (關鍵詞) Classical convolutional neural network en_US dc.subject (關鍵詞) Quantum convolutional neural network en_US dc.subject (關鍵詞) Image classification en_US dc.subject (關鍵詞) Trend prediction en_US dc.title (題名) 基於古典與量子卷積神經網絡的時間序列圖像分析 zh_TW dc.title (題名) Time Series Image Analysis by Classical and Quantum Convolutional Neural Networks en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) [1] Cong, I., Choi, S., & Lukin, M. D. (2019). Quantum Convolutional Neural Networks. Nature Physics, 15, 1273-1278.[2] Eckman, J. P., Kamphorst, S. O., & Ruelle, D. (1987). Recurrence Plots of Dynamical Systems. Europhysics Letters, 4 (91), 973-977.[3] Erhan, D., Bengio, Y., Courville, A., Manzagol, P. A., & Vincent, P. (2010). Why Does Unsupervised Pre-training Help Deep Learning? Journal of Machine Learning Research, 11, 625-660.[4] Heaton, J. B., Polson, N. G., & Witte, J. H. (2016). Deep Learning for Finance: Deep Portfolios. Applied Stochastic Models in Business and Industry, 33, 3-12.[5] Heaton, J. B., Polson, N. G., & Witte, J. H. (2016). Deep Portfolio Theory. arXiv:1605.07230.[6] Kavukcuoglu, K., Sermanet, P., Boureau, Y. L., Gregor, K., Mathieu, M., & LeCun, Y. (2010). Learning Convolutional Feature Hierarchies for Visual Recognition. Neural Information Processing Systems, 1, 1090-1098.[7] Kaye, P., Laflamme, R., & Mosca, M. (2019). An Introduction to Quantum Computing. Oxford University Press.[8] Kerenidis, I., & Prakash, A. (2016). Quantum Recommendation Systems. Innovations in Theoretical Computer Science Conference, 49, 1-21.[9] Kerenidis, I., Landman, J., Luongo, A., & Prakash, A. (2018). Q-means: A Quantum Algorithm for Unsupervised Machine Learning. Neural Information Processing Systems.[10] Kerenidis, I., Landman, J., & Prakash, A. (2020). Quantum Algorithms for Deep Con- volutional Neural Network. International Conference on Learning Representations.[11] Kitaev, A. Y., Shen, A. H., & Vyalyi, M. N. (1999). Classical and Quantum Computa- tion. American Mathematical Society.[12] Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017). ImageNet Classification with Deep Convolutional Neural Networks. Communications of the Association for Computing Machinery, 60(6), 84-90.[13] Lai, C. (2018). Analysis of the predictive ability of time series using convolutional neural network. National Cheng-Chi University.[14] Le, Q. V., Ngiam, J., Chen, Z., Chia, D., Koh, P. W., & Ng, A. Y. (2010). Tiled Convolutional Neural Networks. Neural Information Processing Systems, 1, 1279-1287.[15] LeCun, Y., & Bengio, Y. (1995). Convolutional Networks for Images, Speech, and Time Series. The Handbook of Brain Theory and Neural Networks, 255-258.[16] LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based Learning Ap- plied to Document Recognition. Proceedings of the Institute of Electrical and Electronics Engineers, 86(11), 2278-2324.[17] Martin, T., Hagan, M. T., Demuth, H. B., Beale, M. H., & Jesús, O. D. (2014). Neural Network Design. Martin Hagan.[18] Nakahara, M., & Ohmi, T. (2008). Quantum Ccomputing From Linear Algebra to Physical Realizations. Chemical Rubber Company Press.[19] Nielsen, M. A., & Chuang, I. L. (2010). Quantum Computation and Quantum Infor- mation. University Press of Cambridge.[20] Sakurai, J. J., & Napolitano, J.J. (2014). Modern Quantum Mechanics. Pearson Edu- cation Limited.[21] Scherer, W. (2019). Mathematics of Quantum Computing. Springer Nature Switzerland AG.[22] Shankar, R. (1994). Principles of Quantum Mechanics. Plenum Press.[23] Susskind, L., & Friedman, A. (2014). Quantum Mechanics, The Theoretical Minimum. Perseus Books Group.[24] Wang, Z., & Oates, T. (2015). Imaging Time-Series to Improve Classification and Impu- tation. Proceedings of the International Conference on Artificial Intelligence, 3939-3945.[25] Wu, J. (2017). Introduction to Convolutional Neural Networks. Nanjing University.[26] Zettili, N. (2009). Quantum Mechanics Concepts and Applications. John Wiley & Sons Limited. zh_TW dc.identifier.doi (DOI) 10.6814/NCCU202001059 en_US
