Limits...
Machine-learning techniques for geochemical discrimination of 2011 Tohoku tsunami deposits.

Kuwatani T, Nagata K, Okada M, Watanabe T, Ogawa Y, Komai T, Tsuchiya N - Sci Rep (2014)

Bottom Line: After an exhaustive search of all 262,144 (= 2(18)) combinations of the 18 analysed elements, we observed several tens of combinations with discrimination rates higher than 99.0%.The analytical results show that elements such as Ca and several heavy-metal elements are important for discriminating tsunami deposits from marine sedimentary rocks.The proposed methodology has the potential to aid in the identification of past tsunamis by using other tsunami proxies.

View Article: PubMed Central - PubMed

Affiliation: Graduate School of Environmental Studies, Tohoku University, Sendai 980-8579, Japan.

ABSTRACT
Geochemical discrimination has recently been recognised as a potentially useful proxy for identifying tsunami deposits in addition to classical proxies such as sedimentological and micropalaeontological evidence. However, difficulties remain because it is unclear which elements best discriminate between tsunami and non-tsunami deposits. Herein, we propose a mathematical methodology for the geochemical discrimination of tsunami deposits using machine-learning techniques. The proposed method can determine the appropriate combinations of elements and the precise discrimination plane that best discerns tsunami deposits from non-tsunami deposits in high-dimensional compositional space through the use of data sets of bulk composition that have been categorised as tsunami or non-tsunami sediments. We applied this method to the 2011 Tohoku tsunami and to background marine sedimentary rocks. After an exhaustive search of all 262,144 (= 2(18)) combinations of the 18 analysed elements, we observed several tens of combinations with discrimination rates higher than 99.0%. The analytical results show that elements such as Ca and several heavy-metal elements are important for discriminating tsunami deposits from marine sedimentary rocks. These elements are considered to reflect the formation mechanism and origin of the tsunami deposits. The proposed methodology has the potential to aid in the identification of past tsunamis by using other tsunami proxies.

No MeSH data available.


Related in: MedlinePlus

Determination of the appropriate decision hyperplane.(a) Decision hyperplane in the compositional space for the combination [Mg, Si, Ca] with the highest discrimination rate among the combinations of 7 major elements. (b–d) The scatter diagrams of test data for discrimination using a support vector machine (SVM) for [Mg, Si, Ca] (b), all 18 elements (c), and the optimal 11 elements (d). The discrimination rates are 91.2%, 95.6%, and 100%, respectively. The horizontal axis is Si, and the vertical axis is w·x + b for each combination of elements.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4233330&req=5

f2: Determination of the appropriate decision hyperplane.(a) Decision hyperplane in the compositional space for the combination [Mg, Si, Ca] with the highest discrimination rate among the combinations of 7 major elements. (b–d) The scatter diagrams of test data for discrimination using a support vector machine (SVM) for [Mg, Si, Ca] (b), all 18 elements (c), and the optimal 11 elements (d). The discrimination rates are 91.2%, 95.6%, and 100%, respectively. The horizontal axis is Si, and the vertical axis is w·x + b for each combination of elements.

Mentions: For our first task, we considered the multi-dimensional compositional space defined by the bulk compositions and then determined the decision hyperplane, which divides the samples into tsunami deposits and non-tsunami sediments. This type of problem is known as supervised classification in the fields of machine learning and pattern recognition32. In supervised classification, the available labelled data (those for which the class is known) are regarded as the training problem of the supervisor and are used to determine the decision hyperplane that best classifies the data. In this study, we used N-dimensional bulk compositional data x, labelled as either tsunami or non-tsunami sediment, to determine a linear decision hyperplane in the N-dimensional compositional space: w·x + b = 0, where w is the weight vector that determines the slope of the hyperplane and b is the bias, as shown in Fig. 2a. For this classification, we used a powerful supervised-clustering method, the SVM, which was developed during the 1990s33.


Machine-learning techniques for geochemical discrimination of 2011 Tohoku tsunami deposits.

Kuwatani T, Nagata K, Okada M, Watanabe T, Ogawa Y, Komai T, Tsuchiya N - Sci Rep (2014)

Determination of the appropriate decision hyperplane.(a) Decision hyperplane in the compositional space for the combination [Mg, Si, Ca] with the highest discrimination rate among the combinations of 7 major elements. (b–d) The scatter diagrams of test data for discrimination using a support vector machine (SVM) for [Mg, Si, Ca] (b), all 18 elements (c), and the optimal 11 elements (d). The discrimination rates are 91.2%, 95.6%, and 100%, respectively. The horizontal axis is Si, and the vertical axis is w·x + b for each combination of elements.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4233330&req=5

f2: Determination of the appropriate decision hyperplane.(a) Decision hyperplane in the compositional space for the combination [Mg, Si, Ca] with the highest discrimination rate among the combinations of 7 major elements. (b–d) The scatter diagrams of test data for discrimination using a support vector machine (SVM) for [Mg, Si, Ca] (b), all 18 elements (c), and the optimal 11 elements (d). The discrimination rates are 91.2%, 95.6%, and 100%, respectively. The horizontal axis is Si, and the vertical axis is w·x + b for each combination of elements.
Mentions: For our first task, we considered the multi-dimensional compositional space defined by the bulk compositions and then determined the decision hyperplane, which divides the samples into tsunami deposits and non-tsunami sediments. This type of problem is known as supervised classification in the fields of machine learning and pattern recognition32. In supervised classification, the available labelled data (those for which the class is known) are regarded as the training problem of the supervisor and are used to determine the decision hyperplane that best classifies the data. In this study, we used N-dimensional bulk compositional data x, labelled as either tsunami or non-tsunami sediment, to determine a linear decision hyperplane in the N-dimensional compositional space: w·x + b = 0, where w is the weight vector that determines the slope of the hyperplane and b is the bias, as shown in Fig. 2a. For this classification, we used a powerful supervised-clustering method, the SVM, which was developed during the 1990s33.

Bottom Line: After an exhaustive search of all 262,144 (= 2(18)) combinations of the 18 analysed elements, we observed several tens of combinations with discrimination rates higher than 99.0%.The analytical results show that elements such as Ca and several heavy-metal elements are important for discriminating tsunami deposits from marine sedimentary rocks.The proposed methodology has the potential to aid in the identification of past tsunamis by using other tsunami proxies.

View Article: PubMed Central - PubMed

Affiliation: Graduate School of Environmental Studies, Tohoku University, Sendai 980-8579, Japan.

ABSTRACT
Geochemical discrimination has recently been recognised as a potentially useful proxy for identifying tsunami deposits in addition to classical proxies such as sedimentological and micropalaeontological evidence. However, difficulties remain because it is unclear which elements best discriminate between tsunami and non-tsunami deposits. Herein, we propose a mathematical methodology for the geochemical discrimination of tsunami deposits using machine-learning techniques. The proposed method can determine the appropriate combinations of elements and the precise discrimination plane that best discerns tsunami deposits from non-tsunami deposits in high-dimensional compositional space through the use of data sets of bulk composition that have been categorised as tsunami or non-tsunami sediments. We applied this method to the 2011 Tohoku tsunami and to background marine sedimentary rocks. After an exhaustive search of all 262,144 (= 2(18)) combinations of the 18 analysed elements, we observed several tens of combinations with discrimination rates higher than 99.0%. The analytical results show that elements such as Ca and several heavy-metal elements are important for discriminating tsunami deposits from marine sedimentary rocks. These elements are considered to reflect the formation mechanism and origin of the tsunami deposits. The proposed methodology has the potential to aid in the identification of past tsunamis by using other tsunami proxies.

No MeSH data available.


Related in: MedlinePlus