Using machine learning as an aid to seismic geomorphology, which attributes are the best input?

2019 ◽  
Vol 7 (3) ◽  
pp. SE1-SE18 ◽  
Author(s):  
Lennon Infante-Paez ◽  
Kurt J. Marfurt
Keyword(s):  
Machine Learning ◽  
Volcanic Rocks ◽  
High Amplitude ◽  
Machine Learning Algorithms ◽  
Pyroclastic Flows ◽  
Seismic Survey ◽  
Self Organizing Maps ◽  
Target Pattern ◽  
Seismic Geomorphology ◽  
Amplitude Data

Volcanic rocks with intermediate magma composition indicate distinctive patterns in seismic amplitude data. Depending on the processes by which they were extruded to the surface, these patterns may be chaotic, moderate-amplitude reflectors (indicative of pyroclastic flows) or continuous high-amplitude reflectors (indicative of lava flows). We have identified appropriate seismic attributes that highlight the characteristics of such patterns and use them as input to self-organizing maps to isolate these volcanic facies from their clastic counterpart. Our analysis indicates that such clustering is possible when the patterns are approximately self-similar, such that the appearance of objects does not change at different scales of observation. We adopt a workflow that can help interpreters to decide what methods and what attributes to use as an input for machine learning algorithms, depending on the nature of the target pattern of interest, and we apply it to the Kora 3D seismic survey acquired offshore in the Taranaki Basin, New Zealand. The resulting clusters are then interpreted using the limited well control and principles of seismic geomorphology.

Seismic expression and geomorphology of igneous bodies: A Taranaki Basin, New Zealand, case study

2017 ◽  
Vol 5 (3) ◽  
pp. SK121-SK140 ◽  
Author(s):  
Lennon Infante-Paez ◽  
Kurt J. Marfurt
Keyword(s):  
New Zealand ◽  
Volcanic Rocks ◽  
Pyroclastic Flows ◽  
3D Seismic ◽  
Submarine Volcano ◽  
Self Organizing Maps ◽  
Seismic Geomorphology ◽  
Carbonate Buildups ◽  
Oil Flow

Very little research has been done on volcanic rocks by the oil industry due to the misconception that these rocks cannot be “good reservoirs.” However, in the past two decades, significant quantities of hydrocarbons have been produced from volcanic rocks in China, New Zealand, and Argentina. In frontier basins, volcanic piles are sometimes misinterpreted to be hydrocarbon anomalies and/or carbonate buildups. Unlike clastic and carbonate systems, the 3D seismic geomorphology of igneous systems is only partially documented. We have integrated 3D seismic data, well logs, well reports, core data, and clustering techniques such as self-organizing maps to map two distinct facies (pyroclastic and lava flows), within a Miocene submarine volcano in the Taranaki Basin, New Zealand. Three wells; Kora-1–3 drilled the pyroclastic facies within the volcano encountering evidence of a petroleum system, whereas the Kora-4 well drilled the lava-flow facies, which was barren of hydrocarbons. By integrating results from geochemistry and basin modeling reports prepared for Crown Mineral, New Zealand, we concluded that the reason that Kora-4 was dry was due to a lack of source charge — not to the absence of reservoir quality. Moreover, the Kora-1 well drilled a thick sequence (>[Formula: see text]) of pyroclastic flows in this submarine volcano by chance and found high peaks of gas in the mudlogs near the top 25 m of this sequence. A long-term test in this upper volcanic section resulted in 32 API oil flow of 668 barrels of oil per day for 254 h — a result that challenges the misconception that volcanic rocks cannot be good reservoirs.

scholarly journals Machine Learning-Based Probabilistic Lithofacies Prediction from Conventional Well Logs: A Case from the Umiat Oil Field of Alaska

Energies ◽  
2020 ◽  
Vol 13 (18) ◽  
pp. 4862
Author(s):  
Nilesh Dixit ◽  
Paul McColgan ◽  
Kimberly Kusler
Keyword(s):  
Machine Learning ◽  
Oil Field ◽  
Machine Learning Algorithms ◽  
Self Organizing Maps ◽  
Log Data ◽  
Core Data ◽  
Core Samples ◽  
Resolution Graph ◽  
Reservoir Facies ◽  
Self Organizing

A good understanding of different rock types and their distribution is critical to locate oil and gas accumulations in the subsurface. Traditionally, rock core samples are used to directly determine the exact rock facies and what geological environments might be present. Core samples are often expensive to recover and, therefore, not always available for each well. Wireline logs provide a cheaper alternative to core samples, but they do not distinguish between various rock facies alone. This problem can be overcome by integrating limited core data with largely available wireline log data with machine learning. Here, we presented an application of machine learning in rock facies predictions based on limited core data from the Umiat Oil Field of Alaska. First, we identified five sandstone reservoir facies within the Lower Grandstand Member using core samples and mineralogical data available for the Umiat 18 well. Next, we applied machine learning algorithms (ascendant hierarchical clustering, self-organizing maps, artificial neural network, and multi-resolution graph-based clustering) to available wireline log data to build our models trained with core-driven information. We found that self-organizing maps provided the best result among other techniques for facies predictions. We used the best self-organizing maps scheme for predicting similar reservoir facies in nearby uncored wells—Umiat 23H and SeaBee-1. We validated our facies prediction results for these wells with observed seismic data.

scholarly journals Machine Learning in Discriminating Active Volcanoes of the Hellenic Volcanic Arc

2021 ◽  
Vol 11 (18) ◽  
pp. 8318
Author(s):  
Athanasios G. Ouzounis ◽  
George A. Papakostas
Keyword(s):  
Machine Learning ◽  
Volcanic Rock ◽  
Volcanic Rocks ◽  
Major Elements ◽  
Machine Learning Algorithms ◽  
Geochemical Data ◽  
Geochemical Composition ◽  
Volcanic Arc ◽  
New Approach ◽  
Geochemical Fingerprinting

Identifying the provenance of volcanic rocks can be essential for improving geological maps in the field of geology and providing a tool for the geochemical fingerprinting of ancient artifacts like millstones and anchors in the field of geoarchaeology. This study examines a new approach to this problem by using machine learning algorithms (MLAs). In order to discriminate the four active volcanic regions of the Hellenic Volcanic Arc (HVA) in Southern Greece, MLAs were trained with geochemical data of major elements, acquired from the GEOROC database, of the volcanic rocks of the Hellenic Volcanic Arc (HVA). Ten MLAs were trained with six variations of the same dataset of volcanic rock samples originating from the HVA. The experiments revealed that the Extreme Gradient Boost model achieved the best performance, reaching 93.07% accuracy. The model developed in the framework of this research was used to implement a cloud-based application which is publicly accessible at This application can be used to predict the provenance of a volcanic rock sample, within the area of the HVA, based on its geochemical composition, easily obtained by using the X-ray fluorescence (XRF) technique.

Exhaustive Probabilistic Neural Network for attribute selection and supervised seismic facies classification

2020 ◽  
pp. 1-67
Author(s):  
David Lubo-Robles ◽  
Thang Ha ◽  
Sivaramakrishnan Lakshmivarahan ◽  
Kurt J. Marfurt ◽  
Matthew J. Pranter
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Probabilistic Neural Network ◽  
Learning Algorithms ◽  
Computational Cost ◽  
Component Analysis ◽  
Machine Learning Algorithms ◽  
Seismic Facies ◽  
Seismic Survey ◽  
Facies Classification

Machine learning algorithms such as principal component analysis (PCA), independent component analysis (ICA), self-organizing maps (SOM), and artificial neural networks (ANN), have been used by geoscientists to not only accelerate the interpretation of their data, but also to provide a more quantitative estimate of the likelihood that any voxel belongs to a given facies. Identifying the best combination of attributes needed to perform either supervised or unsupervised machine learning tasks continues to be the most-asked question by interpreters. In the past decades, stepwise regression and genetic algorithms have been used together with supervised learning algorithms to select the best number and combination of attributes. For reasons of computational efficiency, these techniques do not test all the seismic attribute combinations, potentially leading to a suboptimal classification. In this study, we develop an exhaustive probabilistic neural network (PNN) algorithm which exploits the PNN’s capacity in exploring non-linear relationships to obtain the optimal attribute subset that best differentiates target seismic facies of interest. We show the efficacy of our proposed workflow in differentiating salt from non-salt seismic facies in a Eugene Island seismic survey, offshore Louisiana. We find that from seven input candidate attributes, the Exhaustive PNN is capable of removing irrelevant attributes by selecting a smaller subset of four seismic attributes. The enhanced classification using fewer attributes also reduces the computational cost. We then use the resulting facies probability volumes to construct the 3D distribution of the salt diapir geobodies embedded in a stratigraphic matrix.

scholarly journals Evaluation of Shock Detection Algorithm for Road Vehicle Vibration Analysis

Vibration ◽  
2018 ◽  
Vol 1 (2) ◽  
pp. 220-238 ◽  
Author(s):  
Julien Lepine ◽  
Vincent Rouillard
Keyword(s):  
Machine Learning ◽  
Assessment Tool ◽  
High Amplitude ◽  
Detection Algorithm ◽  
Machine Learning Algorithms ◽  
Assessment Tools ◽  
Detection Methods ◽  
Support Vector ◽  
Road Vehicle ◽  
Shock Detection

The ability to characterize shocks which occur during road transport is a vital prerequisite for the design of optimized protective packaging, which can assist in reducing cost and waste related to products and good transport. Many methods have been developed to detect shocks buried in road vehicle vibration signals, but none has yet considered the nonstationary nature of vehicle vibration and how, individually, they fail to accurately detect shocks. Using machine learning, several shock detection methods can be combined, and the reliability and accuracy of shock detection can also be improved. This paper presents how these methods can be integrated into four different machine learning algorithms (Decision Tree, k-Nearest Neighbors, Bagged Ensemble, and Support Vector Machine). The Pseudo-Energy Ratio/Fall-Out (PERFO) curve, a novel classification assessment tool, is also introduced to calibrate the algorithms and compare their detection performance. In the context of shock detection, the PERFO curve has an advantage over classical assessment tools, such as the Receiver Operating Characteristic (ROC) curve, as it gives more importance to high-amplitude shocks.

scholarly journals Machine Learning Platform for Profiling and Forecasting at Microgrid Level

2019 ◽  
Vol 15 (1) ◽  
pp. 21-29
Author(s):  
Enea Mele ◽  
Charalambos Elias ◽  
Aphrodite Ktena
Keyword(s):  
Machine Learning ◽  
Electricity Consumption ◽  
Machine Learning Algorithms ◽  
Small Scale ◽  
Short Time Scale ◽  
Self Organizing Maps ◽  
Learning Platform ◽  
Electricity Use ◽  
Control Technologies ◽  
And Control

AbstractThe shift towards distributed generation and microgrids has renewed the interest in forecasting algorithms and methods, which need to take into account the advances in information, metering and control technologies in order to address the challenges of forecasting problems. Technologies such as machine learning have been proven useful for short-term electricity load forecasting, especially for microgrids, as they can also take into account several types of historical data and can adapt to changes often encountered in small-scale systems and on a short time scale. In this paper, we present a flexible and easily customized modular toolbox, called Divinus, for electricity use profiling and forecasting in microgrids. Divinus may support a variety of machine learning algorithms for forecasting and profiling that can be used independently or combined. For demonstration purposes, we have implemented Self-Organizing Maps for profiling and k-Neighbors for forecasting. The testing of the platform was based on electricity consumption data of the Euripus campus of the National and Kapodistrian University of Athens in Evia, Greece, from January 2010 till March 2018. The tests that have been carried out so far show that the platform can be easily customized and the algorithms examined yield high accuracy and acceptable mean errors for the case of a university campus energy profile.

scholarly journals On Efficiency of Selected Machine Learning Algorithms for Intrusion Detection in Software Defined Networks

2016 ◽  
Vol 62 (3) ◽  
pp. 247-252 ◽  
Author(s):  
Damian Jankowski ◽  
Marek Amanowicz
Keyword(s):  
Machine Learning ◽  
Vector Quantization ◽  
Network Traffic ◽  
Learning Algorithms ◽  
Machine Learning Algorithms ◽  
Software Defined Network ◽  
Software Defined Networks ◽  
Self Organizing Maps ◽  
Machine Learning Methods ◽  
Data Plane

Abstract We propose a concept of using Software Defined Network (SDN) technology and machine learning algorithms for monitoring and detection of malicious activities in the SDN data plane. The statistics and features of network traffic are generated by the native mechanisms of SDN technology. In order to conduct tests and a verification of the concept, it was necessary to obtain a set of network workload test data. We present virtual environment which enables generation of the SDN network traffic. The article examines the efficiency of selected machine learning methods: Self Organizing Maps and Learning Vector Quantization and their enhanced versions. The results are compared with other SDN-based IDS.

Machine learning classifiers for detecting and classifying major explosions and paroxysms at Stromboli volcano (Italy) using radar and optical satellite imagery

2021 ◽  
Author(s):  
Ciro Del Negro ◽  
Claudia Corradino ◽  
Eleonora Amato ◽  
Federica Torrisi ◽  
Sonia Calvari
Keyword(s):  
Machine Learning ◽  
Satellite Imagery ◽  
Google Earth ◽  
Machine Learning Algorithms ◽  
Environmental Data ◽  
Pyroclastic Flows ◽  
Moderate Intensity ◽  
Explosive Activity ◽  
Stromboli Volcano ◽  
Novel Approach

<p>The persistent explosive activity of Stromboli is characterized by several hundred of moderate-intensity events per day. These explosions eject pyroclastic fragments to the height of some tens of meters, which fall a short distance from the summit vents. Occasionally, major explosions eject pyroclastic material to more than a few hundred meters high, which can fall outside the crater terrace on the area visited by tourists. The frequency of these phenomena is variable, with an average of 2 events per year. Paroxysms, violent explosions that produce eruptive columns more than 3 km high and are often associated with pyroclastic flows, can also occur at Stromboli. Ballistic blocks associated with these explosions can reach up to 4 m in diameter and fall on the hinabited areas. Paroxysms are rare (5 events in the last 20 years) and their occurrence frequency varies over time. Nevertheless, major explosions and paroxysms represent the main danger to visitors and inhabitants of the Stromboli Island. Here, we propose a novel approach to detect and classify the type of explosive activity occurring on Stromboli volcano by combining radar and optical satellite imagery with machine learning algorithms. In particular, we considered the plume height, the summit area temperature, and the area affected by large ballistic projectiles as the discriminant factors to distinguish between ordinary activity, major explosions and paroxysms. These factors are retrieved from both radar (Sentinel-1-GRD) and multi-spectral (Landsat-MSI and TIR) satellite images and fed to a machine learning classifier. A retrospective analysis is conducted investigating the main explosive events that have occurred since 1983. This algorithm is based on the in the Google Earth Engine (GEE), which is a cloud computing platform for environmental data analysis from local to planetary scales, with fast access and processing of satellite data from different missions.</p>

scholarly journals Performance Research on different Machine Learning Algorithms for Detection of Sleepy Spindles from EEG signals

2019 ◽  
Vol 8 (9S2) ◽  
pp. 203-208
Keyword(s):  
Machine Learning ◽  
Learning Algorithms ◽  
Transient State ◽  
High Amplitude ◽  
Machine Learning Algorithms ◽  
Computer Interface ◽  
Eeg Signals ◽  
Theta Waves ◽  
Computer Device ◽  
Performance Research

Now a days spindles caused by drowsiness and it has become a very serious issue to accidents. A constant and long driving makes the human brain to a transient state between sleepy and awake. In this BCI plays a major role, where the captured signals from brain neurons are transferred to a computer device. In this paper, I considered the data which are collected from single Electroencephalography (EEG) using Brain Computer Interface (BCI) from the electrodes C3-A1 and C4- A1.Generally these sleepy spindles are present in the theta waves, whose are slower and high amplitude when compared to Alpha and Beta waves and the frequency in ranges from 4 – 8 Hz. The aim of this paper to analyse the accuracy of different machine learning algorithms to identify the spindles.

Sign in / Sign up

Export Citation Format

Share Document