On using kernel integration by graphical LASSO to study partial correlations between heterogeneous data sets

Integration of unstructured and very diverse data is often required for a deeper understanding of complex biological systems. In order to uncover communalities between heterogeneous data, the data is often harmonized by constructing a kernel and numerical integration is performed. In this study we propose a method for data integration in the framework of an undirected graphical model, where the nodes represent individual data sources of varying nature in terms of complexity and underlying distribution, and where the edges represent the partial correlation between two blocks of data. We propose a modified GLASSO for estimation of the graph, with a combination of cross-validation and extended Bayes Information Criterion for sparsity tuning. Furthermore, hierarchical clustering on the weighted consensus kernels from a fixed network is used to partitioning the samples into different classes. Simulations show increasing ability to uncover true edges with increasing sample size and signal to noise. Likewise, identification of non existing edges towards disconnected nodes is feasible. The framework is demonstrated for integration of longitudinal symptom burden data from the 2nd and 3rd year of life with 21 diseases precursors as well as the development of asthma and eczema at the age of 6 years from 403 children from the COPSAC2010 mother-child cohort, suggesting that maternal predisposition as well as being born preterm indirectly lead to higher risk of asthma via increased respiratory symptom burden.

Download Full-text

Ontology-Based Correlation Detection Among Heterogeneous Data Sets: A Case Study of University Campus Issues

2020 IEEE Third International Conference on Artificial Intelligence and Knowledge Engineering (AIKE) ◽

10.1109/aike48582.2020.00014 ◽

2020 ◽

Author(s):

Yuto Tsukagoshi ◽

Shusaku Egami ◽

Yuichi Sei ◽

Yasuyuki Tahara ◽

Akihiko Ohsuga

Keyword(s):

Heterogeneous Data ◽

Data Sets ◽

University Campus

Download Full-text

Comparing global tomography-derived and gravity-based upper mantle density models

Geophysical Journal International ◽

10.1093/gji/ggaa091 ◽

2020 ◽

Vol 221 (3) ◽

pp. 1542-1554 ◽

Cited By ~ 2

Author(s):

B C Root

Keyword(s):

Seismic Tomography ◽

Upper Mantle ◽

Seismic Data ◽

Gravity Data ◽

Heterogeneous Data ◽

Clear Correlation ◽

Data Sets ◽

Satellite Gravity ◽

Similar Density ◽

Similar Peak

SUMMARY Current seismic tomography models show a complex environment underneath the crust, corroborated by high-precision satellite gravity observations. Both data sets are used to independently explore the density structure of the upper mantle. However, combining these two data sets proves to be challenging. The gravity-data has an inherent insensitivity in the radial direction and seismic tomography has a heterogeneous data acquisition, resulting in smoothed tomography models with de-correlation between different models for the mid-to-small wavelength features. Therefore, this study aims to assess and quantify the effect of regularization on a seismic tomography model by exploiting the high lateral sensitivity of gravity data. Seismic tomography models, SL2013sv, SAVANI, SMEAN2 and S40RTS are compared to a gravity-based density model of the upper mantle. In order to obtain similar density solutions compared to the seismic-derived models, the gravity-based model needs to be smoothed with a Gaussian filter. Different smoothening characteristics are observed for the variety of seismic tomography models, relating to the regularization approach in the inversions. Various S40RTS models with similar seismic data but different regularization settings show that the smoothening effect is stronger with increasing regularization. The type of regularization has a dominant effect on the final tomography solution. To reduce the effect of regularization on the tomography models, an enhancement procedure is proposed. This enhancement should be performed within the spectral domain of the actual resolution of the seismic tomography model. The enhanced seismic tomography models show improved spatial correlation with each other and with the gravity-based model. The variation of the density anomalies have similar peak-to-peak magnitudes and clear correlation to geological structures. The resolvement of the spectral misalignment between tomographic models and gravity-based solutions is the first step in the improvement of multidata inversion studies of the upper mantle and benefit from the advantages in both data sets.

Download Full-text

PCA for heterogeneous data sets in a distributed data mining

Proceedings of the Fourth Annual ACM Bangalore Conference on - COMPUTE '11 ◽

10.1145/1980422.1980451 ◽

2011 ◽

Author(s):

E. Chandra ◽

P. Ajitha

Keyword(s):

Data Mining ◽

Heterogeneous Data ◽

Distributed Data Mining ◽

Data Sets ◽

Distributed Data

Download Full-text

Bayesian Uncertainty Estimation for Gaussian Graphical Models and Centrality Indices

10.31234/osf.io/7kude ◽

2021 ◽

Author(s):

Joran Jongerling ◽

Sacha Epskamp ◽

Donald Ray Williams

Keyword(s):

Bayesian Estimation ◽

Graphical Models ◽

Estimation Methods ◽

Centrality Measures ◽

Gaussian Graphical Models ◽

Graphical Lasso ◽

Study Results ◽

Partial Correlations ◽

Good Coverage ◽

Better Than

Gaussian Graphical Models (GGMs) are often estimated using regularized estimation and the graphical LASSO (GLASSO). However, the GLASSO has difficulty estimating(uncertainty in) centrality indices of nodes. Regularized Bayesian estimation might provide a solution, as it is better suited to deal with bias in the sampling distribution ofcentrality indices. This study therefore compares estimation of GGMs with a Bayesian GLASSO- and a Horseshoe prior to estimation using the frequentist GLASSO in an extensive simulation study. Results showed that out of the two Bayesian estimation methods, the Bayesian GLASSO performed best. In addition, the Bayesian GLASSOperformed better than the frequentist GLASSO with respect to bias in edge weights, centrality measures, correlation between estimated and true partial correlations, andspecificity. With respect to sensitivity the frequentist GLASSO performs better.However, sensitivity of the Bayesian GLASSO is close to that of the frequentist GLASSO (except for the smallest N used in the simulations) and tends to be favored over the frequentist GLASSO in terms of F1. With respect to uncertainty in the centrality measures, the Bayesian GLASSO shows good coverage for strength andcloseness centrality. Uncertainty in betweenness centrality is estimated less well, and typically overestimated by the Bayesian GLASSO.

Download Full-text