使用深度卷積神經網路學習進行電阻率斷層掃描 (ERT) 解釋的觀測研究

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：65

、訪客IP：3.144.96.159

姓名

普古婉(Puguh Hiskiawan) 查詢紙本館藏

畢業系所

地球科學學系

論文名稱

使用深度卷積神經網路學習進行電阻率斷層掃描 (ERT) 解釋的觀測研究
(Observational Study of the Electrical Resistivity Tomography (ERT) Interpretation Using Deep Convolutional Neural Network Learning)

相關論文

★ 利用RTL (Region-Time-Length) 演算法探討921 集集大地震之前兆現象	★ 集集餘震b值與碎形維度分析
★ 應用太空大地測量法探討台南地區之地表變形	★ 電容耦合地電阻探測系統應用於地下管線與坑道之研究
★ 以交叉對比分析地震的時空分佈行態	★ 利用PI方法研究地震前兆活動
★ 臺灣深部電性構造及其板塊構造意義	★ 利用Pattern Informatics研究1999年台灣集集與2008年中國汶川地震之前兆現象
★ 模擬地震前兆行為之數值模型	★ 地電法於地下掩埋物調查之研究
★ 利用經驗模態分解法(EMD)探討潮汐效應對地震活動的影響	★ 利用LURR方法探討臺灣1994年後大地震之前兆現象
★ 利用遠距沙堆模型探討特徵地震之準週期性	★ 台灣天然電磁場觀測研究
★ 一維滑塊模型事件叢集特性分析與復發時間統計	★ TCDP井下地震儀之微地震紀錄的特性

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

在本論文中，實施深度卷積神經網絡 (DCNN) 學習，為電阻率層析成像 (ERT) 處理數據提供另一種解釋。地球物理軟件專家對描繪反演電阻率剖面有一個共同的解釋。儘管如此，該解釋對地下的不確定性有一些限制，特別是在確定電阻率分佈的主要質量和顯示揭示的電阻率源方面。因此，觀察性研究旨在對短缺提供適當的解釋。
基於上述問題，本研究開發了幾種合成 ERT 模型，並使用具有局部降水引起的地下擾動的特定野外 ERT 數據集作為第一個觀測工作。在觀察到 [V/I] 處理之後，創建了輸入圖像。它們在 DCNN 學習中用於第二次觀察性研究是有形的。為了構建學習，使用了兩種架構技術：I 形 (ERT-INET) 和 U 形 (ERT-UNET) 。學習過程的最終結果是預測圖像。 Dropout 正則化提高了圖像產量。這些方法用於防止學習中的過度擬合。
成功的 DCNN 呈現了數據集的學習曲線。合成 ERT 模型的曲線結果表明準確度學習在 90% 以上，損失學習在 2% 以下；現場 ERT 揭示了大約 80% 的準確學習和低於 4% 的損失學習。一些統計評估已被用來評估學習表現；結果顯示相似的平均百分比。這項工作導致了預測圖像的開發，該圖像根據在自然地下檢查的類別產生一定程度的分類。
觀察工作的第三階段包括分類評估。接受者操作特徵 (ROC) 曲線用於比較兩種方案之間的預測圖像。該曲線能夠證明兩種架構之間的顯著差異。在三個電阻率類別中，混淆矩陣涵蓋了真實和預測的感知。觀察性研究的結果證明了兩種架構的成就。與這兩種架構一樣，合成 ERT 模型在所有類型的評估中都取得了 85% 以上的成績，包括現場 ERT。
本研究最後一項的觀察工作為綜合觀察，主要在於整合不同剖面產生的結果，也就是預測的反演剖面與實際剖面。結果顯示，ERT-UNET的性能較慢於其他方法。地球物理軟體專家生成反演剖面。一般性結論主要在地下大尺度構造及地球既有異常構造之揭示上發揮作用。這些方法更適用於ERT即時資料集及特定特徵的解釋方面。雖然這些發現可以被視為傳統地球物理方法的替代方案，但並不能替代傳統的地球物理軟體。

摘要(英)

In this dissertation, deep convolutional neural network (DCNN) learning is implemented to provide an alternative interpretation for electrical resistivity tomography (ERT) processing data. Geophysical software experts have a common interpretation for delineating an inversion resistivity profile. Nonetheless, the interpretation has a few limitations on the uncertainty subsurface, particularly in determining the predominant quality of resistivity distributions and displaying a revealed resistivity source. As an outcome, the observational study aims to offer an appropriate interpretation of the shortages.
Based on the aforementioned issues, this study developed several synthetic ERT models and used a specific field ERT dataset with subsurface perturbation due to local precipitation as the first observational work. Following observed [V/I] processing, an input image was created. They are tangible to use in DCNN learning for the second observational study. To construct the learning, two architecture techniques were used: I-shape (ERT-INET) and U-shape (ERT-UNET). The end result of the learning process is a predictive image. Dropout regularization improves image yield. The methods were used to prevent overfitting in the learning process.
The successful DCNN presented a learning curve for the datasets. The curve results for synthetic ERT models indicated an accuracy learning above 90% and a loss learning below 2%; field ERT revealed an accuracy learning around 80% and a loss learning below 4%. Some statistical evaluations have been employed to evaluate learning performance, and the results showed a similar average percentage. This work has resulted in the development of a prediction image that yields a degree of classification based on the categories that were examined in the natural subsurface.
The third phase of observational work included classification assessments. A receiver operating characteristic (ROC) curve was used to compare the predictive image between the two schemes. The curve was able to demonstrate the significant difference between the two architectures. In three resistivity categories, a confusion matrix covers true and predicted perceptions. The findings of the observational study demonstrated the accomplishments of two architectures. The synthetic ERT model, like both architectures, had achieved greater than 85% for all modules of assessments, including field ERT.
The final piece of observational work for this study was a comprehensive observation, which was consolidating among profile yields, that is, prediction with inversion and ground truth profiles. The performance of the ERT-UNET is slower than the others. A geophysical software expert generates the inversion profiles. The general conclusions predicted the great subsurface and revealed the Earth′s existing anomalies. They are better suited for ERT real-time datasets, and the learning interpretations have particular features. While these discoveries can be considered an alternative approach to geophysical methods, the techniques cannot substitute traditional geophysical software.

關鍵字(中)

★ 電阻率斷層掃描 (ERT)
★ 卷積神經網絡 (CNN)
★ 深度學習
★ Wenner-Schlumberger
★ 地球物理方法

關鍵字(英)

★ Electrical Resistivity Tomography (ERT)
★ Convolutional Neural Network (CNN)
★ Deep Learning
★ Wenner-Schlumberger
★ Geophysical Methods

論文目次

摘要 i
ABSTRACT iii
Acknowledgments v
Table of Contents vii
List of Figures ix
List of Tables xi
CHAPTER 1 Introduction 1
1.1 Background 1
1.2 Problem Formulation 3
1.3 Research Objective 4
1.4 Research Contributions 4
1.5 Research Originality 4
CHAPTER 2 Theoretical Basis 5
2.1 Fundamental of Electrical Resistivity Tomography (ERT) 5
2.2 ERT Subsurface Recognition 8
2.3 ERT Method Construction 10
2.4 Deep Learning Techniques 12
2.5 Convolutional Neural Network 12
2.6 Fully Connected Neural Network 17
2.6.1 Perceptron Learning 17
2.6.2 Activation Function 18
2.6.3 ANN Structure 21
2.6.4 Training Mechanisms 21
2.6.5 Parameter Initialization 23
2.6.6 Loss Function 25
2.6.7 Optimizer 25
2.7 Regularization 26
2.7.1 Dropout 27
CHAPTER 3 Research Methodology 29
3.1 Literature Study 29
3.2 Research Procedures 29
3.3 Preparation of Tools and Dataset 30
3.3.1 Apparatus and Equipment Specifications 30
3.3.2 Dataset Compositions 31
3.4 Preprocessing Datasets and data Collections 31
3.4.1 Synthetic ERT Models 31
3.4.1.1 Single rectangular body 33
3.4.1.2 Double up and down rectangular bodies 34
3.4.1.3 Double side by side rectangular bodies 34
3.4.1.4 Double declined adjacent rectangular bodies 35
3.4.1.5 Three rectangular bodies 36
3.4.2 Field ERT Sections 37
CHAPTER 4 General DCNN Architectures 41
4.1 Introduction 41
4.2 Datasets used in the observations 42
4.3 Performance Evaluations of The Learning Dataset 43
4.4 Results and Discussions 44
4.4.1 DCNN for I-shape Architecture (ERT-INET) 44
4.4.1.1 DCNN ERT-INET Performance Evaluations of Synthetics ERT Models 44
4.4.1.2 DCNN ERT-INET Prediction Profiles of Synthetics ERT Models 45
4.4.1.3 DCNN ERT-INET Performance Evaluations of Field ERT Sections 49
4.4.1.4 DCNN ERT-INET Prediction Profiles of Field ERT Sections 50
4.4.2 DCNN for U-shape Architecture (ERT-UNET) 51
4.4.2.1 DCNN ERT-UNET Performance Evaluations of Synthetic ERT Models 51
4.4.2.2 DCNN ERT-UNET Prediction Profiles of Synthetic ERT Models 52
4.4.2.3 DCNN ERT-UNET Performance Evaluations of Field ERT Sections 55
4.4.2.4 DCNN ERT-UNET Prediction Profiles of Field ERT Sections 56
4.5 Summary 57
CHAPTER 5 Performance valuations of DCNN classification 59
5.1 Introduction 59
5.2 Datasets used in the works 60
5.3 Classifier Assessments 60
5.4 Results and Discussions 63
5.4.1 DCNN Classification Yields from I-shape Architecture (ERT-INET) 63
5.4.1.1 Classifier Assessments of the ERT-INET for Synthetic ERT models 63
5.4.1.2 Classifier Assessments of the ERT-INET for Field ERT sections 73
5.4.2 DCNN Classification Yields from U-shape Architecture (ERT-UNET) 75
5.4.2.1 Classifier Assessments of the ERT-UNET for Synthetic ERT models 75
5.4.2.2 Classifier Assessments of the ERT-UNET for Field ERT sections 86
5.5 Summary 88
CHAPTER 6 Consolidate of ERT Interpretation Profile 89
6.1 Introduction 89
6.2 Datasets Used for Consolidations 90
6.3 Model Evaluations of the Comprehensive Observations 90
6.4 Results and Discussions 91
6.4.1 Consolidate to comprehensive observation for Synthetic ERT Models 91
6.4.2 Consolidate to comprehensive observation for Field ERT Sections 96
6.5 Summary 97
CHAPTER 7 Conclusions 99
7.1 Research Conclusions 99
7.2 Research Limitations 100
7.3 Future Research 100
Bibliographies 101
Appendix A. DCNN Features and Source Codes 111
Appendix B. Performance Results of DCNN Learning 117
Appendix C. Discrepancy Compilations of the AUC Values 118
Appendix D. ERT Inversion Profiles 119
Appendix E. Published Article 123

參考文獻

[1] E. A. Robinson and D. Clark, Basic Geophysics. 2017.
[2] J. Milsom and A. Eriksen, Field Geophysics, 4th ed. 2011.
[3] J. . McNeill, Electrical conductivity of soils and rocks, Geonics limited, vol. Technical. p. 21, 1980.
[4] S. Szalai, A. Koppán, & L. Szarka, Effect of positional inaccuracies on multielectrode results, Acta Geod. Geophys. Hungarica, vol. 43, no. 1, pp. 33–42, 2008, doi: 10.1556/AGeod.43.2008.1.3.
[5] M. H. Loke et al., Optimized arrays for 2-D resistivity survey lines with a large number of electrodes, J. Appl. Geophys., vol. 112, pp. 136–146, 2015, doi: 10.1016/j.jappgeo.2014.11.011.
[6] M. H. Loke, Tutorial: 2D and 3D electrical imaging surveys, no. May, 2015, [Online]. Available: http://www.geotomosoft.com/coursenotes.zip.
[7] K. Sudha, M. Israil, S. Mittal, & J. Rai, Soil characterization using electrical resistivity tomography and geotechnical investigations, J. Appl. Geophys., vol. 67, no. 1, pp. 74–79, 2009, doi: 10.1016/j.jappgeo.2008.09.012.
[8] A. Devi et al., Subsurface soil characterization using geoelectrical and geotechnical investigations at a bridge site in Uttarakhand Himalayan region, J. Appl. Geophys., vol. 144, pp. 78–85, 2017, doi: 10.1016/j.jappgeo.2017.07.005.
[9] C. Oberdörster, Hydrological Characterization of a Forest Soil Using Electrical Resistivity Tomography, Univ. Bonn, p. 151, 2010.
[10] A. K. Lysdahl et al., Comparison between 2D and 3D ERT inversion for engineering site investigations - A case study from Oslo Harbour, Near Surf. Geophys., vol. 15, no. 2, pp. 201–209, 2017, doi: 10.3997/1873-0604.2016052.
[11] R. Saad, M. N. M. Nawawi, & E. T. Mohamad, Groundwater detection in alluvium using 2-D electrical resistivity tomography (ERT), Electron. J. Geotech. Eng., vol. 17 D, no. January, pp. 369–376, 2012.
[12] K. Orowe & V. S. Rangarajan, Geo-electrical resistivity and groundwater flow modells for characterization of a hardrock aquifer system, vol. 1, no. 1, pp. 12–31, 2012.
[13] M. Camporese et al., Coupled and uncoupled hydrogeophysical inversions using ensemble Kalman filter assimilation of ERT-monitored tracer test data, Water Resour. Res., vol. 51, no. 5, pp. 3277–3291, 2015, doi: 10.1002/2014WR016017.
[14] A. Casas et al., Assessing aquifer vulnerability to pollutants by electrical resistivity tomography (ERT) at a nitrate vulnerable zone in NE Spain, Environ. Geol., vol. 54, no. 3, pp. 515–520, 2008, doi: 10.1007/s00254-007-0844-1.
[15] C. S. Park, J. H. Jeong, H. W. Park, & K. Kim, Experimental study on electrode method for electrical resistivity survey to detect cavities under road pavements, Sustain., vol. 9, no. 12, 2017, doi: 10.3390/su9122320.
[16] C. Hilbich, C. Fuss, & C. Hauck, Automated time-lapse electrical resistivity tomography ( ERT ) for improved process analysis and long-term monitoring of frozen ground, Permafr. Periglac. Process., vol. 22, no. 4, pp. 306–319, 2010.
[17] M. K. Park et al., Application of electrical resistivity tomography (ERT) technique to detect underground cavities in a karst area of South Korea, Environ. Earth Sci., vol. 71, no. 6, pp. 2797–2806, 2014, doi: 10.1007/s12665-013-2658-7.
[18] D. F. Rucker et al., Electrical resistivity characterization of a reclaimed gold Mine to delineate acid rock drainage pathways, Mine Water Environ., vol. 28, no. 2, pp. 146–157, 2009, doi: 10.1007/s10230-009-0072-x.
[19] P. B. Wilkinson et al., Practical aspects of applied optimized survey design for electrical resistivity tomography, Geophys. J. Int., vol. 189, no. 1, pp. 428–440, 2012, doi: 10.1111/j.1365-246X.2012.05372.x.
[20] A. V. Christiansen & E. Auken, A global measure for depth of investigation, Near Surf. 2010 - 16th Eur. Meet. Environ. Eng. Geophys., vol. 77, no. 4, 2010, doi: 10.3997/2214-4609.20144778.
[21] A. P. Aizebeokhai, 2D and 3D geoelectrical resistivity imaging: Theory and field design, Sci. Res. Essays, vol. 5, no. 23, pp. 3592–3605, 2010.
[22] M. I. I. Abu-Shariah, Determination of cave geometry by using a geoelectrical resistivity inverse model, Eng. Geol., vol. 105, no. 3–4, pp. 239–244, 2009, doi: 10.1016/j.enggeo.2009.02.006.
[23] M. H. Loke, P. B. Wilkinson, et al, Optimized arrays for resistivity measurements confined to the perimeter of a survey area, Near Surf. Geosci. 2015 - 21st Eur. Meet. Environ. Eng. Geophys., no. April 2016, pp. 646–650, 2015, doi: 10.3997/2214-4609.201413793.
[24] M. H. Loke, J. E. Chambers et a.l, Recent developments in the direct-current geoelectrical imaging method, J. Appl. Geophys., vol. 95, pp. 135–156, 2013, doi: 10.1016/j.jappgeo.2013.02.017.
[25] T. Dahlin, The development of DC resistivity imaging techniques, Comput. Geosci., vol. 27, no. 9, pp. 1019–1029, 2001, doi: 10.1016/S0098-3004(00)00160-6.
[26] T. L. Ho, 3-D inversion of borehole-to-surface electrical data using a back-propagation neural network, J. Appl. Geophys., vol. 68, no. 4, pp. 489–499, 2009, doi: 10.1016/j.jappgeo.2008.06.002.
[27] O. L. Johnson & A. P. Aizebeokhai, Application of Artificial Neural Network for the Inversion of Electrical Resistivity Data, J. Informatics Math. sciiences, vol. 9, no. 2, pp. 297–316, 2017.
[28] A. Stanley Raj, Y. Srinivas et al., A novel and generalized approach in the inversion of geoelectrical resistivity data using Artificial Neural Networks (ANN), J. Earth Syst. Sci., vol. 123, no. 2, pp. 395–411, 2014, doi: 10.1007/s12040-014-0402-7.
[29] A. Neyamadpour, S. Taib, et al, Using artificial neural networks to invert 2D DC resistivity imaging data for high resistivity contrast regions: A MATLAB application, Comput. Geosci., vol. 35, no. 11, pp. 2268–2274, 2009, doi: 10.1016/j.cageo.2009.04.004.
[30] B. Liu et al., Deep Learning Inversion of Electrical Resistivity Data, IEEE Trans. Geosci. Remote Sens., vol. 58, no. 8, pp. 5715–5728, 2020, doi: 10.1109/TGRS.2020.2969040.
[31] Z. Geng & Y. Wang, Automated design of a convolutional neural network with multi-scale filters for cost-efficient seismic data classification, Nat. Commun., vol. 11, no. 1, 2020, doi: 10.1038/s41467-020-17123-6.
[32] J. Huang & R. L. Nowack, Machine Learning Using U-Net Convolutional Neural Networks for the Imaging of Sparse Seismic Data, Pure Appl. Geophys., vol. 177, no. 6, pp. 2685–2700, 2020, doi: 10.1007/s00024-019-02412-z.
[33] B. Liu et al., GPRInvNet: Deep Learning-Based Ground-Penetrating Radar Data Inversion for Tunnel Linings, IEEE Trans. Geosci. Remote Sens., 2021, doi: 10.1109/TGRS.2020.3046454.
[34] K. Ishitsuka et al., Object Detection in Ground-Penetrating Radar Images Using a Deep Convolutional Neural Network and Image Set Preparation by Migration, Int. J. Geophys., vol. 2018, 2018, doi: 10.1155/2018/9365184.
[35] C. Schaefer et al., Deep convolutional neural networks as strong gravitational lens detectors, Astron. Astrophys., vol. 611, pp. 1–9, 2018, doi: 10.1051/0004-6361/201731201.
[36] S. Yu et al., Deep learning for gravitational-wave data analysis: A resampling white-box approach, arXiv, vol. 2018, no. July, pp. 5715–5728, 2020, doi: 10.1051/0004-6361/201731201.
[37] S. Vjacheslav & P. Irina, Artificial neural network inversion of magnetotelluric data in terms of three‐dimensional earth macroparameters, Geophys. J. Int., vol. 142, no. 1, pp. 15–26, 2010.
[38] T. Günther & C. Rücker, Electrical Resistivity Tomography (ERT) in geophysical applications-state of the art and future challenges, Researchgate.Net, no. January, pp. 1–4, 2012, _state_of_the_art_and_future_challenges/file/504635151e4096dead.pdf.
[39] S. Szalai, et al., Colinear null arrays in geoelectrics, vol. 7, no. 1, p. 7962, 2005.
[40] M. H. Loke & T. Dahlin, A comparison of the Gauss-Newton and quasi-Newton methods in resistivity imaging inversion, J. Appl. Geophys., vol. 49, no. 3, pp. 149–162, 2002, doi: 10.1016/S0926-9851(01)00106-9.
[41] T. Dahlin & M. H. Loke, Resolution of 2D Wenner resistivity imaging as assessed by numerical modelling, J. Appl. Geophys., vol. 38, no. 4, pp. 237–249, 1998, doi: 10.1016/S0926-9851(97)00030-X.
[42] A. Binley & A. Kemna, DC Resistivity and Induced Polarization Methods, no. October 2016. 2005.
[43] A. Samouëlian et al., Electrical resistivity survey in soil science: A review, Soil Tillage Res., vol. 83, no. 2, pp. 173–193, 2005, doi: 10.1016/j.still.2004.10.004.
[44] M. H. Loke, P. B. Wilkinson, & J. E. Chambers, Fast computation of optimized electrode arrays for 2D resistivity surveys, Comput. Geosci., vol. 36, no. 11, pp. 1414–1426, 2010, doi: 10.1016/j.cageo.2010.03.016.
[45] J. & D. K. Hurwitz, Machine Learning for Dummies. IBM Limited Edition, 2018.
[46] T. Perol, M. Gharbi, & M. Denolle, Convolutional neural network for earthquake detection and location, Sci. Adv., vol. 4, no. 2, pp. 2–9, 2018, doi: 10.1126/sciadv.1700578.
[47] I. Goodfellow, Y. Bengio, & A. Courville, Deep Learning. MIT Press, 2016.
[48] M. Alaslani & L. Elrefaei, Convolutional Neural Network Based Feature Extraction for IRIS Recognition, Int. J. Comput. Sci. Inf. Technol., vol. 10, pp. 65–78, 2018, doi: 10.5121/ijcsit.2018.10206.
[49] Q. Zhao & Z. Shang, Deep learning and Its Development, J. Phys. Conf. Ser., vol. 1948, no. 1, 2021, doi: 10.1088/1742-6596/1948/1/012023.
[50] K. Simonyan & A. Zisserman, Very deep convolutional networks for large-scale image recognition, 3rd Int. Conf. Learn. Represent. ICLR 2015 - Conf. Track Proc., pp. 1–14, 2015.
[51] C. Szegedy et al., Going deeper with convolutions, Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., vol. 07-12-June, pp. 1–9, 2015, doi: 10.1109/CVPR.2015.7298594.
[52] M. Reichstein et al., Deep learning and process understanding for data-driven Earth system science, Nature, vol. 566, no. 7743, pp. 195–204, 2019, doi: 10.1038/s41586-019-0912-1.
[53] A. Vedaldi & K. Lenc, MatConvNet: Convolutional neural networks for MATLAB, MM 2015 - Proc. 2015 ACM Multimed. Conf., pp. 689–692, 2015, doi: 10.1145/2733373.2807412.
[54] C. Dong et al., Image Super-Resolution Using Deep Convolutional Networks, IEEE Trans. Pattern Anal. Mach. Intell., vol. 38, no. 2, pp. 295–307, 2016, doi: 10.1109/TPAMI.2015.2439281.
[55] L. Fausett, Fundamentals of Neural Network: Architectures, Algorithms, and Applications, vol. C–18, no. 6. USA: Prentice-Hall, Inc., 1994.
[56] M. Van Der Baan & C. Jutten, Neural networks in geophysical applications, Geophysics, vol. 65, no. 4, pp. 1032–1047, 2000, doi: 10.1190/1.1444797.
[57] D. Meyer, F. Leisch, & K. Hornik, The support vector machine under test, Neurocomputing, vol. 55, no. 1–2, pp. 169–186, 2003, doi: 10.1016/S0925-2312(03)00431-4.
[58] R. Jie, J. Gao, A. Vasnev, & M. N. Tran, Regularized flexible activation function combination for deep neural networks, Proc. - Int. Conf. Pattern Recognit., vol. abs/2007.1, pp. 2001–2008, 2020, doi: 10.1109/ICPR48806.2021.9412370.
[59] Y. LeCun et al., Backpropagation applied to digit recognition, Neural computation, vol. 1, no. 4. pp. 541–551, 1989,
[60] A. L. Maas, A. Y. Hannun, & A. Y. Ng, Rectifier nonlinearities improve neural network acoustic models, ICML Work. Deep Learn. Audio, Speech Lang. Process., vol. 28, 2013.
[61] Y. N. Dauphin et al, Identifying and attacking the saddle point problem in high-dimensional non-convex optimization, Adv. Neural Inf. Process. Syst., vol. 4, no. January, pp. 2933–2941, 2014.
[62] S. Ruder, An overview of gradient descent optimization algorithms, pp. 1–14, 2016,.
[63] P. Rajendra Kumar & E. B. K. Manash, Deep learning: A branch of machine learning, J. Phys. Conf. Ser., vol. 1228, no. 1, 2019, doi: 10.1088/1742-6596/1228/1/012045.
[64] V. Dumoulin & F. Visin, A guide to convolution arithmetic for deep learning, pp. 1–31, 2016,
[65] E. Shelhamer, J. Long, & T. Darrell, Fully Convolutional Networks for Semantic Segmentation, IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, no. 4, pp. 640–651, 2017, doi: 10.1109/TPAMI.2016.2572683.
[66] Y. Lecun, Y. Bengio, & G. Hinton, Deep learning, Nature, vol. 521, no. 7553, pp. 436–444, 2015, doi: 10.1038/nature14539.
[67] T. Chai & R. R. Draxler, Root mean square error (RMSE) or mean absolute error (MAE)? -Arguments against avoiding RMSE in the literature, Geosci. Model Dev., vol. 7, no. 3, pp. 1247–1250, 2014, doi: 10.5194/gmd-7-1247-2014.
[68] H. Kuwajima, Backpropagation in Convolutional Neural Network, 2014. .
[69] T. Perol, M. Gharbi, & M. Denolle, Convolutional neural network for earthquake detection and location, Sci. Adv., vol. 4, no. 2, pp. 1–9, 2018, doi: 10.1126/sciadv.1700578.
[70] K. Zhang et al., Beyond a Gaussian denoiser: Residual learning of deep CNN for image denoising, IEEE Trans. Image Process., vol. 26, no. 7, pp. 3142–3155, 2017, doi: 10.1109/TIP.2017.2662206.
[71] S. Mahdizadehaghdam, A. Panahi, & H. Krim, Sparse generative adversarial network, Proc. - 2019 Int. Conf. Comput. Vis. Work. ICCVW 2019, pp. 3063–3071, 2019, doi: 10.1109/ICCVW.2019.00369.
[72] G. Cybenko, Approximation by superpositions of a sigmoidal function, Math. Control. Signals, Syst., vol. 2, no. 4, pp. 303–314, 1989, doi: 10.1007/BF02551274.
[73] A. Mosca & G. D. Magoulas, Adapting Resilient Propagation for Deep Learning, 2015,
[74] O. OZTURK, et al, Comparison of Fully Convolutional Networks (FCN) and U-Net for Road Segmentation from High Resolution Imageries, Int. J. Environ. Geoinformatics, vol. 7, no. 3, pp. 272–279, 2020, doi: 10.30897/ijegeo.737993.
[75] Z. Fang et al., Integration of convolutional neural network and conventional machine learning classifiers for landslide susceptibility mapping, Comput. Geosci., vol. 139, no. October 2019, p. 104470, 2020, doi: 10.1016/j.cageo.2020.104470.
[76] D. P. Kingma & J. L. Ba, Adam: A method for stochastic optimization, 3rd Int. Conf. Learn. Represent. ICLR 2015 - Conf. Track Proc., no. July, p. 5265, 2015.
[77] E. Cuevas & J. Galvez, An optimization algorithm guided by a machine learning approach, Int. J. Mach. Learn. Cybern., vol. 10, no. 11, pp. 2963–2991, 2019, doi: 10.1007/s13042-018-00915-0.
[78] S. Yu & J. Ma, Data-driven geophysics: From dictionary learning to deep learning, arXiv, pp. 1–73, 2020.
[79] A. Graves, A. R. Mohamed, & G. Hinton, Speech recognition with deep recurrent neural networks, ICASSP, IEEE Int. Conf. Acoust. Speech Signal Process. - Proc., no. 3, pp. 6645–6649, 2013, doi: 10.1109/ICASSP.2013.6638947.
[80] N. Srivastava et al., Dropout: A Simple Way to Prevent Neural Networks from Overfitting, J. Mach. Learn. Res., vol. 15, no. 56, pp. 1929–1958, 2014.
[81] M. Araya-Polo et al., Deep-learning tomography, Lead. Edge, vol. 37, no. 1, pp. 58–66, 2018, doi: 10.1190/tle37010058.1.
[82] B. B. Bougher, Machine learning applications to geophysical data analysis, Physics (College. Park. Md)., no. August, p. 68, 2016, doi: 10.14288/1.0308786.
[83] D. Conway et al., Inverting magnetotelluric responses in a three-dimensional earth using fast forward approximations based on artificial neural networks, Comput. Geosci., vol. 127, no. March, pp. 44–52, 2019, doi: 10.1016/j.cageo.2019.03.002.
[84] D. J. Lary et al., Machine Learning Applications for Earth Observation. 2018.
[85] M. Shahriari et al., A deep learning approach to the inversion of borehole resistivity measurements, Comput. Geosci., vol. 24, no. 3, pp. 971–994, 2020, doi: 10.1007/s10596-019-09859-y.
[86] G. N. Loginov & A. M. Petrov, Automatic detection of geoelectric boundaries according to lateral logging sounding data by applying a deep convolutional neural network, Russ. Geol. Geophys., vol. 60, no. 11, pp. 1319–1325, 2019, doi: 10.15372/RGG2019134.
[87] B. Xu et al., Empirical Evaluation of Rectified Activations in Convolutional Network, 2015,
[88] B. Russell, Machine learning and geophysical inversion - A numerical study, Lead. Edge, vol. 38, no. 7, pp. 512–519, 2019, doi: 10.1190/tle38070512.1.
[89] A. Mariette& K. Rahul, Efficient Learning Machines, vol. 111, no. 479. 1965.
[90] R. Johansson, Numerical python: Scientific computing and data science applications with numpy, SciPy and matplotlib, Second edition. 2018.
[91] F. Bao, X. He, & F. Zhao, Applying Data Mining to the Geosciences Data, Phys. Procedia, vol. 33, pp. 685–689, 2012, doi: 10.1016/j.phpro.2012.05.121.
[92] T. Dahlin & B. Zhou, A numerical comparison of 2D resistivity imaging with 10 electrode arrays, Geophys. Prospect., vol. 52, no. 5, pp. 379–398, 2004, doi: 10.1111/j.1365-2478.2004.00423.x.
[93] Y. Zhu et al., Converting tabular data into images for deep learning with convolutional neural networks, Sci. Rep., vol. 11, no. 1, pp. 1–11, 2021, doi: 10.1038/s41598-021-90923-y.
[94] Z.-S. Chen, Z.-Y. Hseu, & C.-C. Tsai, The Soils of Taiwan. 2015.
[95] S. Y. Liu, Y. J. Du, L. H. Han, & M. F. Gu, Experimental study on the electrical resistivity of soil-cement admixtures, Environ. Geol., vol. 54, no. 6, pp. 1227–1233, 2008, doi: 10.1007/s00254-007-0905-5.
[96] S. Arridge et al., Mini-Workshop: Deep Learning and Inverse Problems, Oberwolfach Reports, vol. 15, no. 1, pp. 559–589, 2019, doi: 10.4171/owr/2018/11.
[97] S. Reddy, K. T. Reddy, & V. Kumari, Optimization of Deep Learning using Various Optimizers , Loss Functions and Dropout, 2019.
[98] R. Donida Labati et at., A novel pore extraction method for heterogeneous fingerprint images using Convolutional Neural Networks, Pattern Recognit. Lett., vol. 113, pp. 58–66, 2018, doi: 10.1016/j.patrec.2017.04.001.
[99] V. Badrinarayanan, A. Kendall, & R. Cipolla, SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation, IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, no. 12, pp. 2481–2495, 2017, doi: 10.1109/TPAMI.2016.2644615.
[100] J. H. Kim et al., Four-dimensional inversion of resistivity monitoring data through Lp norm minimizations, Geophys. J. Int., vol. 195, no. 3, pp. 1640–1656, 2013, doi: 10.1093/gji/ggt324.
[101] B. J. Singstad & C. Tronstad, Convolutional Neural Network and Rule-Based Algorithms for Classifying 12-lead ECGs, Comput. Cardiol. (2010)., vol. 2020-Septe, pp. 2–5, 2020, doi: 10.22489/CinC.2020.227.
[102] P. F. Zhao et al., Electrical imaging of plant root zone: A review, Comput. Electron. Agric., vol. 167, no. June, p. 105058, 2019, doi: 10.1016/j.compag.2019.105058.
[103] S. Cardimona, Electrical Resistivity Techniques for Subsurface Investigation . Department of Geology and Geophysics, University of Missouri-Rolla, Rolla, MO., Dep. Geol. Geophys. Univ. Missouri-Rolla, Roll. MO, no. 1976, pp. 11–0, 2002.
[104] M. H. Loke & R. D. Barker, Least-squares deconvolution of apparent resistivity pseudosections, Geophysics, vol. 60, no. 6, pp. 1682–1690, 1995, doi: 10.1190/1.1443900.
[105] M. H. Loke, H. Kiflu et al., Optimized arrays for 2D resistivity surveys with combined surface and buried arrays, Near Surf. Geophys., vol. 13, no. 5, pp. 505–517, 2015, doi: 10.3997/1873-0604.2015038.
[106] P. Stummer, H. Maurer, & A. G. Green, Experimental design: Electrical resistivity data sets that provide optimum subsurface information, Geophysics, vol. 69, no. 1, pp. 120–139, 2004, doi: 10.1190/1.1649381.
[107] W. Luo, Y. Li, R. Urtasun, & R. Zemel, Understanding the effective receptive field in deep convolutional neural networks, Adv. Neural Inf. Process. Syst., no. Nips, pp. 4905–4913, 2016.
[108] S. Maiti et al., ANN based inversion of DC resistivity data for groundwater exploration in hard rock terrain of western Maharashtra (India), J. Hydrol., vol. 464–465, pp. 294–308, 2012, doi: 10.1016/j.jhydrol.2012.07.020.
[109] P. Krähenbühl & V. Koltun, Efficient inference in fully connected crfs with Gaussian edge potentials, Adv. Neural Inf. Process. Syst. 24 25th Annu. Conf. Neural Inf. Process. Syst. 2011, NIPS 2011, pp. 1–9, 2011.
[110] G. Programming, D. E. Goldberg, & J. R. Koza, Adaptive Learning of Polynomial Networks. 2006.
[111] J. Giraud et al., Impact of uncertain geology in constrained geophysical inversion, ASEG Ext. Abstr., vol. 2018, no. 1, pp. 1–6, 2018, doi: 10.1071/aseg2018abm1_2f.
[112] S. Szalai, A. Novák, & L. Szarka, Depth of investigation and vertical resolution of surface geoelectric arrays, J. Environ. Eng. Geophys., vol. 14, no. 1, pp. 15–23, 2009, doi: 10.2113/JEEG14.1.15.
[113] S. Szalai & L. Szarka, Expanding the possibilities of two-dimensional multielectrode systems, with consideration to earlier geoelectric arrays, J. Appl. Geophys., vol. 75, no. 1, pp. 1–8, 2011, doi: 10.1016/j.jappgeo.2011.06.020.
[114] T. Fawcett, An introduction to ROC analysis, Pattern Recognit. Lett., vol. 27, no. 8, pp. 861–874, 2006, doi: 10.1016/j.patrec.2005.10.010.
[115] K. M. Ting, Confusion Matrix, Encycl. Mach. Learn. Data Min., no. October, pp. 260–260, 2017, doi: 10.1007/978-1-4899-7687-1_50.
[116] M. Studer et al., Discrepancy Analysis of State Sequences, Sociol. Methods Res., vol. 40, no. 3, pp. 471–510, 2011, doi: 10.1177/0049124111415372.
[117] V. Sofia, Confusion Matrix-based Feature Selection Sofia Visa, ConfusionMatrix-based Featur. Sel. Sofia, vol. 710, no. January, p. 8, 2011.
[118] I. Markoulidakis et al., “ulticlass Confusion Matrix Reduction Method and Its Application on Net Promoter Score Classification Problem, Technologies, vol. 9, no. 4, 2021, doi: 10.3390/technologies9040081.
[119] M. Grandini, E. Bagli, & G. Visani, Metrics for Multi-Class Classification: an Overview, pp. 1–17, 2020,
[120] D. Chicco & G. Jurman, The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation, BMC Genomics, vol. 21, no. 1, pp. 1–13, 2020, doi: 10.1186/s12864-019-6413-7.
[121] Y. Taigman, M. Yang, M. Ranzato, & L. Wolf, DeepFace: Closing the gap to human-level performance in face verification, Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., pp. 1701–1708, 2014, doi: 10.1109/CVPR.2014.220.
[122] M. O. Cimpoiaşu, O. Kuras, T. Pridmore,& S. J. Mooney, Potential of geoelectrical methods to monitor root zone processes and structure: A review, Geoderma, vol. 365, no. January, 2020, doi: 10.1016/j.geoderma.2020.114232.
[123] A. Furman, T. P. A. Ferre, & A. W. Warrick, Optimization of ERT Surveys for Monitoring Transient Hydrological Events Using Perturbation Sensitivity and Genetic Algorithms, Vadose Zo. J., vol. 3, no. 4, pp. 1230–1239, 2004, doi: 10.2113/3.4.1230.
[124] A. DEY & H. F. MORRISON, Resistivity Modelling for Arbitrarily Shaped Two‐Dimensional Structures, Geophysical Prospecting, vol. 27, no. 1. pp. 106–136, 1979, doi: 10.1111/j.1365-2478.1979.tb00961.x.
[125] M. Blome, H. Maurer, & S. Greenhalgh, Geoelectric experimental design - Efficient acquisition and exploitation of complete pole-bipole data sets, Geophysics, vol. 76, no. 1, 2011, doi: 10.1190/1.3511350.
[126] Advanced Geosciences (AGI), EarthImager 2D, no. 512, p. 139, 2009.
[127] Y. Y. . Al-najjar & D. C. . Soong, Comparison of Image Quality Assessment: PSNR, HVS, SSIM, UIQI, Int. J. Sci. Eng. Res., vol. 3, no. 8, pp. 1–5, 2012,
[128] S. Furui et al., Fundamental technologies in modern speech recognition, IEEE Signal Process. Mag., vol. 29, no. 6, pp. 16–17, 2012, doi: 10.1109/MSP.2012.2209906.
[129] R. Pires de Lima et al., Deep convolutional neural networks as a geological image classification tool, Sediment. Rec., vol. 17, no. 2, pp. 4–9, 2019, doi: 10.2110/sedred.2019.2.4.
[130] V. Sangeetha & K. J. R. Prasad, Syntheses of novel derivatives of 2-acetylfuro[2,3-a]carbazoles, benzo[1,2-b]-1,4-thiazepino[2,3-a]carbazoles and 1-acetyloxycarbazole-2- carbaldehydes, Indian J. Chem. - Sect. B Org. Med. Chem., vol. 45, no. 8, pp. 1951–1954, 2006, doi: 10.1002/chin.200650130.
[131] Q. Zhang, Q. Yuan, C. Zeng, X. Li, & Y. Wei, Missing data reconstruction in remote sensing image with a unified spatial-temporal-spectral deep convolutional neural network, IEEE Trans. Geosci. Remote Sens., vol. 56, no. 8, pp. 4274–4288, 2018, doi: 10.1109/TGRS.2018.2810208.

指導教授

陳建志(Chien-Chih Chen)

審核日期

2024-1-15

推文