Unsupervised learning on scientific ocean drilling datasets from the South China Sea

doi:10.1007/s11707-018-0704-1

Front. Earth Sci.

2019, Vol. 13

Issue (1) : 180-190 https://doi.org/10.1007/s11707-018-0704-1

RESEARCH ARTICLE

Unsupervised learning on scientific ocean drilling datasets from the South China Sea

Kevin C. TSE¹(

), Hon-Chim CHIU², Man-Yin TSANG³, Yiliang LI¹, Edmund Y. LAM⁴

¹. Department of Earth Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China
². Department of Geography and Centre for Geo-computation Studies, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China
³. Department of Earth Sciences, University of Toronto, Toronto, ON M5S 2M8, Canada
⁴. Department of Electrical and Electronic Engineering, The University of Hong Kong, Pokfulam, Hong Kong, China

Download: PDF(1039 KB) HTML
Export: BibTeX | EndNote | Reference Manager | ProCite | RefWorks

Abstract

Unsupervised learning methods were applied to explore data patterns in multivariate geophysical datasets collected from ocean floor sediment core samples coming from scientific ocean drilling in the South China Sea. Compared to studies on similar datasets, but using supervised learning methods which are designed to make predictions based on sample training data, unsupervised learning methods require no a priori information and focus only on the input data. In this study, popular unsupervised learning methods including K-means, self-organizing maps, hierarchical clustering and random forest were coupled with different distance metrics to form exploratory data clusters. The resulting data clusters were externally validated with lithologic units and geologic time scales assigned to the datasets by conventional methods. Compact and connected data clusters displayed varying degrees of correspondence with existing classification by lithologic units and geologic time scales. K-means and self-organizing maps were observed to perform better with lithologic units while random forest corresponded best with geologic time scales. This study sets a pioneering example of how unsupervised machine learning methods can be used as an automatic processing tool for the increasingly high volume of scientific ocean drilling data.

Keywords machine learning unsupervised learning ODP IODP clustering

Corresponding Author(s): Kevin C. TSE

Just Accepted Date: 24 April 2018 Online First Date: 01 June 2018 Issue Date: 25 January 2019

Cite this article:

Kevin C. TSE,Hon-Chim CHIU,Man-Yin TSANG, et al. Unsupervised learning on scientific ocean drilling datasets from the South China Sea[J]. Front. Earth Sci., 2019, 13(1): 180-190.

URL:

https://academic.hep.com.cn/fesci/EN/10.1007/s11707-018-0704-1
https://academic.hep.com.cn/fesci/EN/Y2019/V13/I1/180

Fig.1 Map showing locations of the four ocean drill sites used in this study.

Tab.1 Summary of the datasets of the four ODP/IODP sites

Tab.2 Summary of input datasets in this study

Class	v₁	v₂	...	vC	Sums
u₁	n11	n12	...	n1C	n1·
u₂	n21	n22	...	n2C	n2·
?	?	?	$?$	?	?
u_R	nR1	nR2	...	nRC	nR·
Sums	n·1	n·2	...	n·C	n_·· = n

Tab.3 Notation for comparing two partitions (^{Hubert and Arabie, 1985}), also referred as contingency table

Tab.4 Validation of clustering results for the four SCS sites with lithological units

Tab.5 Validation of clustering results for the four SCS sites with geological time scales

Tab.6 Clustering results for individual datasets for lithological units

Tab.7 Clustering results for individual datasets for geological time scales

Normalization	Index	1146	1148	U1431	U1433
Original	RI₁	0.65	0.769	0.668	0.525
Original	RI₂	0.233	0.21	0.098	0.097
$x ′ = x − x ‾ σ x$	RI₁	0.829	0.808	0.826	0.697
$x ′ = x − x ‾ σ x$	RI₂	0.568	0.272	0.329	0.236
$x ′ = x − min ? (x) max ? (x) − min ? (x)$	RI₁	0.828	0.846	0.82	0.622
$x ′ = x − min ? (x) max ? (x) − min ? (x)$	RI₂	0.567	0.403	0.3	0.168
$x ′ = ln ? (x − min ? (x) + 1)$	RI₁	0.817	0.859	0.815	0.636
$x ′ = ln ? (x − min ? (x) + 1)$	RI₂	0.551	0.455	0.267	0.214

Tab.8 Clustering results for different normalization methods on lithologic unit

Fig.2 Unsupervised clustering results compared with lithological units and geological time scales for the ODP sites 1146 and 1148. In calculating the RI, the number of clusters are set to equal to the number of lithological units or number of geological time scales. (a) 1146 K-means lith. units (cluster=4, RI₁=0.832, RI₂=0.584); (b) 1146 K-means geo. units (cluster=5, RI₁=0.840, RI₂=0.554); (c) 1148 SOMs lith. units (cluster=7, RI₁=0.869, RI₂=0.503); (d) 1148 RF geo. units (cluster=6, RI₁=0.861, RI₂=0.435).

Fig.3 Unsupervised clustering results compared with lithological units and geological time scales for the IODP sites U1431 and U1433. In calculating the RI, the number of clusters is set to equal the number of lithological units or number of geological time scales. (a) U1431 RF lith. units (cluster=9, RI₁=0.839, RI₂=0.357); (b) U1431 RF geo. units (cluster=5, RI₁=0.731, RI₂=0.425); (c) U1433 SOMs lith. units (cluster=5, RI₁=0.697, RI₂=0.282); (d) U1433 RF geo. units (cluster=4, RI₁=0.706, RI₂=0.254).

1	E WAugustijn, RZurita-Milla (2013). Self-organizing maps as an approach to exploring spatiotemporal diffusion patterns. Int J Health Geogr, 12(1): 60 https://doi.org/10.1186/1476-072X-12-60
2	JBaarsch, M Celebi (2012). Investigation of internal validity measures for k-means clustering. In: Proceedings of the International Multi Conference of Engineers and Computer Scientists
3	EBedini (2009). Mapping lithology of the Sarfartoq carbonatite complex, southern West Greenland, using HyMap imaging spectrometer data. Remote Sens Environ, 113(6): 1208–1219 https://doi.org/10.1016/j.rse.2009.02.007
4	EBedini (2012). Mapping alteration minerals at Malmbjerg molybdenum deposit, central East Greenland, by Kohonen self-organizing maps and matched filter analysis of HyMap data. Int J Remote Sens, 33(4): 939–961 https://doi.org/10.1080/01431161.2010.542202
5	DBenaouda, G Wadge, R BWhitmarsh, R GRothwell, CMacLeod (1999). Inferring the lithology of borehole rocks by applying neural network classifiers to downhole logs: an example from the ocean drilling program. Geophys J Int, 136(2): 477–491 https://doi.org/10.1046/j.1365-246X.1999.00746.x
6	F PBierlein, S J Fraser, W Brown, TLees (2008). Advanced methodologies for the analysis of databases of mineral deposits and major faults. Aust J Earth Sci, 55(1): 79–99 https://doi.org/10.1080/08120090701581406
6	LBreiman, (1984). Classification and Regression Trees. New York: Chapman & Hall, 87–91 https://doi.org/10.1080/08120090701581406
7	LBreiman (2001). Random forests. Mach Learn, 45(1): 5–32 https://doi.org/10.1023/A:1010933404324
8	C DCantrell (2000). Modern Mathematical Methods for Physicists and Engineers. Cambridge University Press
9	SChauhan, W Ruhaak, FKhan, FEnzmann, PMielke, MKersten, ISass (2016). Processing of rock core microtomogrpahy images: using seven different machine learning algorithms. Comput Geosci, 86: 120–128 https://doi.org/10.1016/j.cageo.2015.10.013
10	M JCracknell, A MReading, A WMcNeill (2014). Mapping geology and volcanic-hosted massive sulfide alteration in the Hellyer-Mt Charter region, Tasmania, using Random Forest and Self-Organising Maps. Aust J Earth Sci, 61(2): 287–304 https://doi.org/10.1080/08120099.2014.858081
11	J NGoetz, A Brenning, HPetschko, PLeopold (2015). Evaluating machine learning and statistical prediction techniques for landslide susceptibility modeling. Computers & Geosciences, 81: 1–11
12	MHalkidi, Y Batistakis, MVazirgiannis (2002). Clustering validity checking methods: part II. ACM SIGMOD Rec, 31(3): 19–27
13	LHamel (2009). Knowledge Discovery with Support Vector Machines. New York: John Wiley and Sons, 89–132
14	CHennig (2015). What are the true clusters? Pattern Recognit Lett, 64: 53–62 https://doi.org/10.1016/j.patrec.2015.04.009
15	LHubert, P Arabie (1985). Comparing partitions. J Classif, 2(1): 193–218 https://doi.org/10.1007/BF01908075
16	T LInsua, L Hamel, KMoran, L MAnderson, J MWebster (2015). Advanced classification of carbonate sediments based on physical properties. Sedimentology, 62(2): 590–606 https://doi.org/10.1111/sed.12168
17	JJeong, E Park (2016). Comparative Application of Various Machine Learning Techniques for Lithology Predictions. J Soil Groundw Environ, 21(3): 21–34 https://doi.org/10.7857/JSGE.2016.21.3.021
18	R IKabacoff (2015). R in Action- Data analysis and graphics with R.Greenwich, CT: Manning, 102–112
19	TKohonen (1982). Self-organized formation of topologically correct feature maps. Biol Cybern, 43(1): 59–69 https://doi.org/10.1007/BF00337288
20	TKohonen (2001). Self-Organizing Maps (3rd ed). New York: Springer, 132–154
21	E FKrause (1987). Taxicab Geometry- An Adventure in Non-Euclidean Geometry. Stroud, UK: Dover, 120–132
22	D JLary, A H Alavi, A H Gandomi, A L Walker (2016). Machine learning in geosciences and remote sensing. Geoscience Frontiers, 7(1): 3–10 https://doi.org/10.1016/j.gsf.2015.07.003
23	C FLi, J Lin, D KKulhanek (2014). IODP expedition 349 preliminary report, South China Sea tectonics–Opening of the South China Sea and its implications for Southeast Asian tectonics, climates and deep mantle processes since the late Mesozoic. Technical report
25	GLongo, M Brescia, SDjorgovski, SCavuoti, CDonalek (2014). Data driven discovery in astrophysics. Proceedings of ESA-ESRIN Conference: Big Data from Space 2014, Frascati, Italy
26	JMacQueen (1967). Some methods for classification and analysis of multivariate observations. In: Le Cam L M, Neyman J, eds. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. University of California, 281–297
27	G AMarzo, T L Roush, A Blanco, SFonti, VOrofino (2006). Cluster analysis of planetary remote sensing spectral data. Journal of Geophysical Research, 111: E03002
28	GMoore, A Taira, AKlaus, KBecker, MSaffer, EScreaton (2001). Proc. ODP, Init. Repts., 190. College Station, TX (Ocean Drilling Program)
29	K PMurphy (2012). Machine Learning A Probabilistic Perspective.Cambridge: The MIT Press, 578–490
30	LPeeters, F Bação, VLobo, ADassargues (2007). Exploratory data analysis and clustering of multivariate spatial hydrogeological data by means of GEO3DSOM, a variant of Kohonen’s self-organizing map. Hydrol Earth Syst Sci, 11(4): 1309–1321 https://doi.org/10.5194/hess-11-1309-2007
31	B SPenn (2005). Using self-organizing maps to visualize high-dimensional data. Comput Geosci, 31(5): 531–544 https://doi.org/10.1016/j.cageo.2004.10.009
32	B TPham, D T Bui, I Prakash (2017a). Landslide susceptibility assessment using bagging ensemble based alternating decision trees, logistic regression and J48 decision trees methods: a comparative study. Geotech Geol Eng, 35(6): 2597–2611
33	B TPham, K Khosravi, IPrakash (2017b). Application and comparison of decision tree-based machine learning methods in landside susceptibility assessment at Pauri Garhwal area, Uttarakhand, India. Environmental Processes, 2017, 4(3): 711–730
34	B TPham, D Tien Bui, H VPham, H QLe, IPrakash, M BDholakia (2016). Landslide hazard assessment using random subspace fuzzy rules based classifier ensemble and probability analysis of rainfall data: a case study at Mu Cang Chai District, Yen Bai Province (Viet Nam). Journal of the Indian Society of Remote Sensing, 45: 673–683
35	W MRand (1971). Objective criteria for the evaluation of clustering methods. Journal of the American Statistical Association, 66(336): 846–850
36	B DRipley (1996). Pattern Recognition and Neural Networks. Cambridge University Press, 248–290
37	TRomary, F Ors, JRivoirard, JDeraisme (2015). Unsupervised classification of multivariate geostatistical data: two algorithms. Comput Geosci, 85: 96–103 https://doi.org/10.1016/j.cageo.2015.05.019
38	J LSchnase, T J Lee, C A Mattmann, C S Lynnes, L Cinquini, P MRamirez, A FHart, D NWilliams, DWaliser, PRinsland, W PWebster, D QDuffy, M AMcInerney, G STamkin, G LPotter, LCarriere (2016). Big data challenges in climate science. IEEE Geosciences and Remote Sensing, 4(3): 10–22 https://doi.org/10.1109/MGRS.2015.2514192
39	MTempl, P Filzmoser, CReimann (2008). Cluster analysis applied to regional geochemical data: problems and possibilities. Appl Geochem, 23(8): 2198–2213 https://doi.org/10.1016/j.apgeochem.2008.03.004
40	P XWang, Q Y Li (2009). The South China Sea Paleoceanography and Sedimentology. New York: Springer, 388–421
24	TWarren Liao (2005). Clustering of time series data- a survey. Pattern Recognit, 38(11): 1857–1874 https://doi.org/10.1016/j.patcog.2005.01.025
41	M JWay, J D Scargle, K M Ali, A N Srivastava (2012). Advances in Machine Learning and Data Mining for Astronomy. New York: CRC Press, 240–312
42	RWehrens, L M C Buydens (2007). Self- and super-organising maps in R: the Kohonen package. Journal of Statistical Software, 21(5):1–19
43	YXiong, R Zuo (2016). Recognition of geochemical anomalies using a deep autoencoder network. Comput Geosci, 86: 75–82 https://doi.org/10.1016/j.cageo.2015.10.006
44	XYao, L G Tham, F C Dai (2008). Landslide susceptibility mapping based on Support Vector Machine: a case study on natural slopes of Hong Kong, China. Geomorphology, 101(4): 572–582 https://doi.org/10.1016/j.geomorph.2008.02.011

[1]	Kevin C. Tse, Hon-Chim Chiu, Man-Yin Tsang, Yiliang Li, Edmund Y. Lam. An unsupervised learning approach to study synchroneity of past events in the South China Sea[J]. Front. Earth Sci., 2019, 13(3): 628-640.
[2]	Feng MAO,Minhe JI,Ting LIU. Mining spatiotemporal patterns of urban dwellers from taxi trajectory data[J]. Front. Earth Sci., 2016, 10(2): 205-221.
[3]	Nana SHI, Jinyan ZHAN, Feng WU, Jifu DU. Identification of the core ecosystem services and their spatial heterogeneity in Poyang Lake area[J]. Front Earth Sci Chin, 2009, 3(2): 214-220.

Viewed

Full text

Abstract

Cited

Shared

Discussed