Abstract
Traditional machine learning requires big data as the basis, and it is usually difficult to obtain the desired results in areas where there is a lack of data. This study proposes an innovative transfer learning method to establish a seismic landslide susceptibility evaluation model. The process is as follows: (1) a total of 13 influence factors were selected and combined with landslide points triggered by the Wenchuan and Jiuzhaigou earthquakes to form a big dataset and a small sample dataset, respectively; (2) Artificial Neural Network (ANN) was used to train the big dataset and prepare a pre-training model; (3) the Jiuzhaigou seismic landslide susceptibility evaluation model based on transfer learning was established by using the pre-training model. To reflect the advantages of the transfer learning method more intuitively, this study not only tested the accuracy of the evaluation model but also used ANN to train another evaluation model based on the small sample datasets. And the model accuracy was compared with that of the previous model. The results showed that the frequency ratio (FR) accuracy of the model obtained by transfer learning was higher than that of the model directly trained on a small sample dataset. Additionally, the area under curve (AUC) of the model directly trained on a small sample dataset was only 0.84, whereas the AUC of the model obtained by transfer learning was close to 0.90. The study shows that this method can solve the problems associated with traditional machine learning methods when establishing a seismic landslide susceptibility evaluation model.
Similar content being viewed by others
References
Aditian A, Kubota T, Shinohara Y (2018) Comparison of GIS-based landslide susceptibilitymodels using frequency ratio, logistic regression, and artificial neural network in a tertiary region of Ambon, Indonesia. Geomophology 318:101–111
Akin O, Sayil N (2013) Geophysical survey on a landslide area is in the Trabzon City Centre Northern Turkey, 7th Congress of Balkan Geophysical Society, 1–4, Tiran, Albenia, 7–10 October 2013
Akin O, Sayil N (2016) Site characterization using surface wave methods in the Arsin-Trabzon Province, NE Turkey. Environ Earth Sci 75(72):1–17
Akin O (2020) Investigation of soil characteristics of landscaped areas in south Trabzon-Ortahisar district by active and passive surface wave methods, PhD Thesis, Karadeniz Technical University, Institute of Science, Trabzon, Turkey
Aleotti P, Chowdhury R (1999) Landslide hazard assessment: summery review and new perspective. Bull Eng Geol Env 58(1):28–44
Beven KJ, Kirkby MJ (1979) A physically based, variable contributing area model of basin hydrology. Hydrol Sci Bull 24:43–69
Chang Z, Du Z, Zhang F, Huang F, Chen JW, Li W, Guo Z (2020) Landslide susceptibility prediction based on remote sensing images and gis: comparisons of supervised and unsupervised machine learning models. Remote Sens-BASEL 12(3):502
Dhanya J, Raghukanth STG (2020) Neural network-based hybrid ground motion prediction equations for Western Himalayas and North-Eastern India. Acta Geophys 68:303–324
Duan M, Wang GP, Niu CY (2018) Method of small sample size image recognition based on convolution neural network. Computer Engineering and Design 39(1):224–229
Fang M, Yin J, Zhu XQ (2013) Transfer learning across networks for collective classification. Proc. of the 2013 IEEE 13th Int’l Conf. on Data Mining. 161–170
Fawcett T (2005) An introduction to ROC analysis. Pattern Recogn Lett 27(8):861–874
Gu ZJ, Yang WJ, Zhou JX (2020) Small sample DGA malicious domain names detection method based on transfer learning. Computer Engineering and Applications 1–8
Harmouzi H, Nefeslioglu HA, Rouai M, Sezer EA, Dekayir A, Geokceoglu C (2019) Landslide susceptibility mapping of the Mediterranean coastal zone of Morocco between Oued Laou and El Jebha using artificial neural networks (ANN). Arab J Geosci 12(696):1–18
Huang F, Yao C, Liu W, Li Y, Liu X (2018) Landslide susceptibility assessment in the Nantian area of China: a comparison of frequency ratio model and support vector machine. Geomat Nat Haz Risk 9(1):919–938
Luzi L, Pergalani F (1999) Slope instability in static and dynamic conditions for urban planning: the ‘“Oltre Po Pavese”’ case history (Region Lombardia-Italy). Nat Hazards 20:57–82
Nefeslioglu H, Gokceoglu C, Sonmez H (2008) An assessment on the use of logistic regression and artificial neural networks with different sampling strategies for the preparation of landslide susceptibility maps. Eng Geol 97:171
Pachauri AK, Pant M (1992) Landslide hazard mapping based on geological attributes. Eng Geol 32:81–100
Pan SJ, Yang Q (2010) A survey on transfer learning. Transactions on Knowledge and Data Engineering IEEE 22(10):1345–1359
Pazzi V, Tanteri L, Bicocchi G, D'Ambrosio M, Caselli A, Fanti R (2016) H/V measurements as an effective tool for the reliable detection of landslide slip surfaces: case studies of Castagnola (La Spezia, Italy) and Roccalbegna (Grosseto, Italy), Phys Chem Earth, Parts A/B/C
Pazzi V, Tanteri L, Bicocchi G, D'Ambrosio M, Caselli A, Fanti R (2017) H/V Technique for the rapid detection of landslide slip surface(s): assessment of the optimized measurements spatial distribution, 4th World Landslide Forum, Ljubljana, SLOVENIA, 335–343, May, 2017
Pan C, Huang J, Hao JG, Gong JX, Zhang ZJ (2020) Survey of Semi-supervised Learning Including Few-shot and Zero-shot Learning. Systems Engineering and Electronics 42(10):2246-2256.
Pradhan B, Lee S, Buchroithner M (2010) A GIS-based back-propagation neural network model and its cross-application and validation for landslide susceptibility analyses. Comput Environ Urban 34:216–235
Saha AK, Gupta RP, Arora MK (2002) GIS-based landslide hazard zonation in the Bhagirathi (Ganga) valley. Himalayas Int J Remote Sens 23(2):357–369
Samanta S, Tirumarai SA, Das S (2013) Cross-domain clustering performed by transfer of knowledge across domains. Computer Vision, Pattern Recognition, Image Processing & Graphics. IEEE
Thiery Y, Malet JP, Sterlacchini S, Puissant A, Maquaire O (2007) Landslide susceptibility assessment by bivariate methods at large scales: application to a complex mountainous environment. Geomorphology 92(1):38–59
Tian X, Tao D, Rui Y (2012) Sparse transfer learning for interactive video search reranking. ACM Transactions on Multimedia Computing, Communications, and Applications 8(3):26:1–19
Tom F (2006) An introduction to ROC analysis. Pattern Recogn Lett 27(8):861–874
Uromeihy A, Mahdavifar MR (2000) Landslide hazard zonation of the Khorshrostam area. Iran Bull Eng Geol Environ 58(3):207–213
Wei FM, Zhang JP, Chu Y, Yang J (2014) FSFP: Transfer learning from long texts to the short. Appl Math Inform Sci 8(4):2033–2044
Xu C (2012) Detailed inventory of landslides triggered by the 2008 Wenchuan earthquake and its comparison with other earthquake events in the world. Science & Technology Review 30(25):18–26
Xu C, Xu XW, Wu XY, Dai FC, Yao X, Yao Q (2013) Detailed catalog of landslides triggered by the 2008 Wenchuan earthquake and statistical analysis of their spatial distribution. J Eng Geol 21(1):1–20
Xu C, Xu XW (2014) Construction of basic earthquake-triggered landslides dataset for several large earthquake events at the beginning of the twenty-first century. Seismology and Geology 36(1):90–104
Xu C, Dai FC, Yao X, Zhao Z, Xiao JZ (2010) GIS platform and certainty factor analysis method based Wenchuan earthquake-induced landslide susceptibility evaluation. J Eng Geol 18(1):15–26
Xu C, Wang SY, Xu XW, Zhang H, Tian YY, Ma SY, Fang LH, Lu RQ, Chen LC, Tan XB (2018) A Panorama of Landslides Triggered by the 8 August 2017 Jjiuzhaigou, Sichuan Ms7.0 Earthquake. Seismology and Geology 40(1):232-260.
Zhang Y, Xia HB (2018) Sentiment analysis method based on pre-training convolutional neural networks by distant supervision. Comput Eng Appl 54(13):27–33
Author information
Authors and Affiliations
Corresponding author
Additional information
Highlights
• Proposing an innovative transfer learning method to establish a seismic landslide susceptibility evaluation model of the target area
• Using the frequency ratio (FR) and ROC curve were used to compare the accuracy of the models to reflect the advantages of transfer learning
• Introducing an optimization method to replace the traditional learning rate in the process of this model training, and the results show that the above optimization method can avoid the overfitting phenomenon and improve the training efficiency of the model
Rights and permissions
About this article
Cite this article
Ai, X., Sun, B. & Chen, X. Construction of small sample seismic landslide susceptibility evaluation model based on Transfer Learning: a case study of Jiuzhaigou earthquake. Bull Eng Geol Environ 81, 116 (2022). https://doi.org/10.1007/s10064-022-02601-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10064-022-02601-6