scholarly journals Petrophysically and geologically guided multi-physics inversion using a dynamic Gaussian mixture model

2020 ◽  
Vol 224 (1) ◽  
pp. 40-68 ◽  
Author(s):  
Thibaut Astic ◽  
Lindsey J Heagy ◽  
Douglas W Oldenburg
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Gaussian Mixture ◽  
Quantitative Information ◽  
Geophysical Data ◽  
Magnetic Data ◽  
Data Sets ◽  
Data Set ◽  
Geophysical Surveys ◽  
Geological Information

SUMMARY In a previous paper, we introduced a framework for carrying out petrophysically and geologically guided geophysical inversions. In that framework, petrophysical and geological information is modelled with a Gaussian mixture model (GMM). In the inversion, the GMM serves as a prior for the geophysical model. The formulation and applications were confined to problems in which a single physical property model was sought, and a single geophysical data set was available. In this paper, we extend that framework to jointly invert multiple geophysical data sets that depend on multiple physical properties. The petrophysical and geological information is used to couple geophysical surveys that, otherwise, rely on independent physics. This requires advancements in two areas. First, an extension from a univariate to a multivariate analysis of the petrophysical data, and their inclusion within the inverse problem, is necessary. Secondly, we address the practical issues of simultaneously inverting data from multiple surveys and finding a solution that acceptably reproduces each one, along with the petrophysical and geological information. To illustrate the efficacy of our approach and the advantages of carrying out multi-physics inversions coupled with petrophysical and geological information, we invert synthetic gravity and magnetic data associated with a kimberlite deposit. The kimberlite pipe contains two distinct facies embedded in a host rock. Inverting the data sets individually, even with petrophysical information, leads to a binary geological model: background or undetermined kimberlite. A multi-physics inversion, with petrophysical information, differentiates between the two main kimberlite facies of the pipe. Through this example, we also highlight the capabilities of our framework to work with interpretive geological assumptions when minimal quantitative information is available. In those cases, the dynamic updates of the GMM allow us to perform multi-physics inversions by learning a petrophysical model.

Joint inversion of potential-fields data over the DO-27 kimberlite pipe using a Gaussian mixture model prior

2020 ◽  
Vol 8 (4) ◽  
pp. SS47-SS62
Author(s):  
Thibaut Astic ◽  
Dominique Fournier ◽  
Douglas W. Oldenburg
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Joint Inversion ◽  
Gravity Data ◽  
Physical Property ◽  
Gaussian Mixture ◽  
Kimberlite Pipe ◽  
Magnetic Data ◽  
High Porosity ◽  
Rock Types

We have carried out petrophysically and geologically guided inversions (PGIs) to jointly invert airborne and ground-based gravity data and airborne magnetic data to recover a quasi-geology model of the DO-27 kimberlite pipe in the Tli Kwi Cho (also referred to as TKC) cluster. DO-27 is composed of three main kimberlite rock types in contact with each other and embedded in a granitic host rock covered by a thin layer of glacial till. The pyroclastic kimberlite (PK), which is diamondiferous, and the volcanoclastic kimberlite (VK) have anomalously low density, due to their high porosity, and weak magnetic susceptibility. They are indistinguishable from each other based upon their potential-field responses. The hypabyssal kimberlite (HK), which is not diamondiferous, has been identified as highly magnetic and remanent. Quantitative petrophysical signatures for each rock unit are obtained from sample measurements, such as the increasing density of the PK/VK unit with depth and the remanent magnetization of the HK unit, and are represented as a Gaussian mixture model (GMM). This GMM guides the PGI toward generating a 3D quasi-geology model with physical properties that satisfies the geophysical data sets and the petrophysical signatures. Density and magnetization models recovered individually yield volumes that have physical property combinations that do not conform to any known petrophysical characteristics of the rocks in the area. A multiphysics PGI addresses this problem by using the GMM as a coupling term, but it puts a volume of the PK/VK unit at a location that is incompatible with geologic information from drillholes. To conform to that geologic knowledge, a fourth unit is introduced, PK-minor, which is petrophysically and geographically distinct from the main PK/VK unit. This inversion produces a quasi-geology model that presents good structural locations of the diamondiferous PK unit and can be used to provide a resource estimate or decide the locations of future drillholes.

A framework for petrophysically and geologically guided geophysical inversion using a dynamic Gaussian mixture model prior

2019 ◽  
Vol 219 (3) ◽  
pp. 1989-2012 ◽  
Author(s):  
Thibaut Astic ◽  
Douglas W Oldenburg
Keyword(s):  
Objective Function ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Optimization Problems ◽  
Reference Model ◽  
Gaussian Mixture ◽  
Geological Data ◽  
Geophysical Inversion ◽  
Geophysical Model ◽  
Geological Information

SUMMARY We propose a new framework for incorporating petrophysical and geological information into voxel-based geophysical inversion. By developing the geophysical inverse problem from a probabilistic perspective, we redesign the objective function and the iteration steps as a suite of cyclic optimization problems in which three separate MAP optimization problems are solved using geophysical, petrophysical and geological data, respectively. By quantitatively linking these data into a single framework, we recover a final inverted model that reproduces the observed, or desired, petrophysical and geological features while fitting the geophysical data. To achieve our goal we replace the Gaussian prior, used in the Tikhonov inversion approach, by a Gaussian mixture model. After each geophysical model update, the mixture parameters (means, variances and proportions) are determined by the geophysical model and the expected characteristics of the lithologies through another optimization process using the expectation–maximization algorithm. We then classify the model cells into rock units according to the petrophysical and geological information. These two additional steps over the petrophysical and geological data result in a dynamic update of the reference model and associated weights and guide the inversion towards reproducing the expected petrophysical and geological characteristics. The resulting geophysical objective function does not require extra terms to include the additional petrophysical and geological information; this is an important distinction between our work and previous frameworks that carry out joint geophysical and petrophysical data inversion. We highlight different capabilities of our methodology by inverting magnetotelluric and direct-current resistivity data in 1-D and 2-D, respectively. Finally, we apply our framework to inverting airborne frequency domain data, acquired in Australia, for the detection and characterization of saline contamination of freshwater.

Rapid multi-scale analysis of near-surface geophysical anomaly maps: Application to an archaeo-geophysical data set

Geophysics ◽  
2020 ◽  
pp. 1-41 ◽  
Author(s):  
Jens Tronicke ◽  
Niklas Allroggen ◽  
Felix Biermann ◽  
Florian Fanselow ◽  
Julien Guillemoteau ◽  
...  
Keyword(s):  
Random Noise ◽  
Geophysical Data ◽  
Magnetic Data ◽  
Engineering Practice ◽  
Data Sets ◽  
Stripe Pattern ◽  
Data Set ◽  
Near Surface ◽  
Multi Scale ◽  
Topographic Data

In near-surface geophysics, ground-based mapping surveys are routinely employed in a variety of applications including those from archaeology, civil engineering, hydrology, and soil science. The resulting geophysical anomaly maps of, for example, magnetic or electrical parameters are usually interpreted to laterally delineate subsurface structures such as those related to the remains of past human activities, subsurface utilities and other installations, hydrological properties, or different soil types. To ease the interpretation of such data sets, we propose a multi-scale processing, analysis, and visualization strategy. Our approach relies on a discrete redundant wavelet transform (RWT) implemented using cubic-spline filters and the à trous algorithm, which allows to efficiently compute a multi-scale decomposition of 2D data using a series of 1D convolutions. The basic idea of the approach is presented using a synthetic test image, while our archaeo-geophysical case study from North-East Germany demonstrates its potential to analyze and process rather typical geophysical anomaly maps including magnetic and topographic data. Our vertical-gradient magnetic data show amplitude variations over several orders of magnitude, complex anomaly patterns at various spatial scales, and typical noise patterns, while our topographic data show a distinct hill structure superimposed by a microtopographic stripe pattern and random noise. Our results demonstrate that the RWT approach is capable to successfully separate these components and that selected wavelet planes can be scaled and combined so that the reconstructed images allow for a detailed, multi-scale structural interpretation also using integrated visualizations of magnetic and topographic data. Because our analysis approach is straightforward to implement without laborious parameter testing and tuning, computationally efficient, and easily adaptable to other geophysical data sets, we believe that it can help to rapidly analyze and interpret different geophysical mapping data collected to address a variety of near-surface applications from engineering practice and research.

scholarly journals Clustering algorithms subjected to K-mean and gaussian mixture model on multidimensional data set

2019 ◽  
Vol 7 (2) ◽  
pp. 448 ◽  
Author(s):  
Saadaldeen Rashid Ahmed Ahmed ◽  
Israa Al Barazanchi ◽  
Zahraa A. Jaaz ◽  
Haider Rasheed Abdulshaheed
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Clustering Algorithms ◽  
Gaussian Mixture ◽  
Multidimensional Data ◽  
Data Set ◽  
Multidimensional Data Set

scholarly journals Efficient Moving Vehicle Detection Algorithm for Various Traffic Conditions

2019 ◽  
Vol 8 (3) ◽  
pp. 6069-6076
Keyword(s):  
Real Time ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Gaussian Mixture ◽  
Traffic Monitoring ◽  
Visual Surveillance ◽  
Detection Accuracy ◽  
Data Set ◽  
Moving Vehicle ◽  
Detection Techniques

Many computer vision applications needs to detect moving object from an input video sequences. The main applications of this are traffic monitoring, visual surveillance, people tracking and security etc. Among these, traffic monitoring is one of the most difficult tasks in real time video processing. Many algorithms are introduced to monitor traffic accurately. But most of the cases, the detection accuracy is very less and the detection time is higher which makes the algorithms are not suitable for real time applications. In this paper, a new technique to detect moving vehicle efficiently using Modified Gaussian Mixture Model and Modified Blob Detection techniques is proposed. The modified Gaussian Mixture model generates the background from overall probability of the complete data set and by calculating the required step size from the frame differences. The modified Blob Analysis is then used to classify proper moving objects. The simulation results shows that the method accurately detect the target

scholarly journals AGN dichotomy beyond radio loudness: a Gaussian mixture model analysis

2020 ◽  
Vol 497 (2) ◽  
pp. 1463-1474
Author(s):  
Pedro P B Beaklini ◽  
Allan V C Quadros ◽  
Marcio G B de Avellar ◽  
Maria L L Dantas ◽  
André L F Cançado
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Flux Ratio ◽  
Gaussian Mixture ◽  
Data Sets ◽  
Stellar Objects ◽  
Quasi Stellar Objects ◽  
Different Populations ◽  
5 Ghz ◽  
Two Populations

ABSTRACT Since the discovery of quasi-stellar objects (QSOs), also known as quasars, they have been traditionally subdivided as radio-loud and radio-quiet sources. Whether such division is a misleading effect from a highly heterogeneous single population of objects, or real has yet to be answered. Such dichotomy has been evidenced by observations of the flux ratio between the optical and radio emissions (usually B band and 5 GHz). Evidence of two populations in quasars and samples of a wide diversity of active galactic nuclei (AGNs) has been accumulated over the years. Other quantities beyond radio loudness also seem to show the signature of the existence of two different populations of AGNs. To verify the existence of a dichotomy through different parameters, we employed a soft clustering scheme, based on the Gaussian mixture model (GMM), to classify these objects simultaneously using the following parameters: black hole mass, colour, and R loudness index, as well as the usual radio and B-band luminosity. To investigate whether different kinds of AGNs manifest any population dichotomy, we applied GMM to four independent catalogues composed of both optical and radio information. Our results indicate the persistence of a dichotomy in all data sets, although the discriminating power differs for different choices of parameters. Although the radio loudness parameter alone does not seem to be enough to display the dichotomy, the evidence of two populations of AGNs could persist even if we consider other parameters. Our research suggests that the dichotomy is not a misleading effect but real.

scholarly journals A Machine Learning Recommender System Based on Collaborative Filtering Using Gaussian Mixture Model Clustering

2020 ◽  
Author(s):  
Delshad Fakoor ◽  
Vafa Maihami ◽  
Reza Maihami
Keyword(s):  
Machine Learning ◽  
Recommender Systems ◽  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Recommender System ◽  
Pearson Correlation ◽  
Gaussian Mixture ◽  
Similarity Criteria ◽  
Data Set ◽  
Mixture Model Clustering

Changing and moving toward online shopping has made it necessary to customize customers’ needs and provide them more selective options. The buyers search the products’ features before deciding to purchase items. The recommender systems facilitate the searching task for customers via narrowing down the search space within the specific products that align the customer needs. Clustering, as a typical machine learning approach, is applied in recommender systems. As an information filtering method, a recommender system clusters user’s data to indicate the required factors for more accurate predictions by calculating the similarity between members of a cluster. In this study, using the Gaussian mixture model clustering and considering the scores distance and the value of scores in the Pearson correlation coefficient, a new method is introduced for predicting scores in machine learning recommender systems. To study the proposed method’s performance, a Movie Lens data set is evaluated, and the results are compared to some other recommender systems, including the Pearson correlation coefficients similarity criteria, K-means, and fuzzy C-means algorithms. The simulation results indicate that our method has less error than others by increasing the number of neighbors. The results also illustrate that when the number of users increases, the proposed method’s accuracy will increase. The reason is that the Gaussian mixture clustering chooses similar users and considers the scores distance in choosing similar neighbors to the active user.

scholarly journals POLARIMETRIC SAR DATA GMM CLASSIFICATION BASED ON IMPROVED FREEMAN INCOHERENT DECOMPOSITION

2016 ◽  
Vol XLI-B7 ◽  
pp. 341-345
Author(s):  
S. Rouabah ◽  
M. Ouarzeddine ◽  
B. Azmedroub
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Urban Areas ◽  
Classification Problem ◽  
Gaussian Mixture ◽  
Polarimetric Sar ◽  
Data Set ◽  
Feature Vectors ◽  
Sar Data

Due to the increasing volume of available SAR Data, powerful classification processings are needed to interpret the images. GMM (Gaussian Mixture Model) is widely used to model distributions. In most applications, GMM algorithm is directly applied on raw SAR data, its disadvantage is that forest and urban areas are classified with the same label and gives problems in interpretation. In this paper, a combination between the improved Freeman decomposition and GMM classification is proposed. The improved Freeman decomposition powers are used as feature vectors for GMM classification. The E-SAR polarimetric image acquired over Oberpfaffenhofen in Germany is used as data set. The result shows that the proposed combination can solve the standard GMM classification problem.

scholarly journals Research on dynamic routing algorithm based on gaussian mixture model

2022 ◽  
Vol 355 ◽  
pp. 03017
Author(s):  
Yuzhan Huang
Keyword(s):  
Gaussian Mixture Model ◽  
Mixture Model ◽  
Routing Algorithm ◽  
Gaussian Mixture ◽  
Dynamic Routing ◽  
Distribution Functions ◽  
Environmental Sound ◽  
Data Set ◽  
Original Algorithm ◽  
Sound Detection

In this paper, based on the method of environmental sound detection, a neural network model based on capsule network and Gaussian mixture model is proposed. The model proposed in this paper mainly aims at the disadvantages of dynamic routing algorithm in the capsule network, and proposes a dynamic routing algorithm based on Gaussian mixture model. The improved dynamic routing algorithm assumes that the characteristics of the data conform to the multi-dimensional Gaussian distribution, so the model can learn the distribution of data features by building distribution functions of different classes. The information entropy is used as the activation value of the salient degree of the feature. Through experiments, the accuracy of the proposed algorithm on Urbansound8K data set is more than 92%, which is 4.8% higher than the original algorithm.

Sign in / Sign up

Export Citation Format

Share Document