-
Nonnegative Matrix Factorization for identification of unknown number of sources emitting delayed signals
Authors:
Filip L. Iliev,
Valentin G. Stanev,
Velimir V. Vesselinov,
Boian S. Alexandrov
Abstract:
Factor analysis is broadly used as a powerful unsupervised machine learning tool for reconstruction of hidden features in recorded mixtures of signals. In the case of a linear approximation, the mixtures can be decomposed by a variety of model-free Blind Source Separation (BSS) algorithms. Most of the available BSS algorithms consider an instantaneous mixing of signals, while the case when the mix…
▽ More
Factor analysis is broadly used as a powerful unsupervised machine learning tool for reconstruction of hidden features in recorded mixtures of signals. In the case of a linear approximation, the mixtures can be decomposed by a variety of model-free Blind Source Separation (BSS) algorithms. Most of the available BSS algorithms consider an instantaneous mixing of signals, while the case when the mixtures are linear combinations of signals with delays is less explored. Especially difficult is the case when the number of sources of the signals with delays is unknown and has to be determined from the data as well. To address this problem, in this paper, we present a new method based on Nonnegative Matrix Factorization (NMF) that is capable of identifying: (a) the unknown number of the sources, (b) the delays and speed of propagation of the signals, and (c) the locations of the sources. Our method can be used to decompose records of mixtures of signals with delays emitted by an unknown number of sources in a nondispersive medium, based only on recorded data. This is the case, for example, when electromagnetic signals from multiple antennas are received asynchronously; or mixtures of acoustic or seismic signals recorded by sensors located at different positions; or when a shift in frequency is induced by the Doppler effect. By applying our method to synthetic datasets, we demonstrate its ability to identify the unknown number of sources as well as the waveforms, the delays, and the strengths of the signals. Using Bayesian analysis, we also evaluate estimation uncertainties and identify the region of likelihood where the positions of the sources can be found.
△ Less
Submitted 23 March, 2018; v1 submitted 12 December, 2016;
originally announced December 2016.
-
Identification of release sources in advection-diffusion system by machine learning combined with Green function inverse method
Authors:
Valentin G. Stanev,
Filip L. Iliev,
Scott Hansen,
Velimir V. Vesselinov,
Boian S. Alexandrov
Abstract:
The identification of sources of advection-diffusion transport is based usually on solving complex ill-posed inverse models against the available state- variable data records. However, if there are several sources with different locations and strengths, the data records represent mixtures rather than the separate influences of the original sources. Importantly, the number of these original release…
▽ More
The identification of sources of advection-diffusion transport is based usually on solving complex ill-posed inverse models against the available state- variable data records. However, if there are several sources with different locations and strengths, the data records represent mixtures rather than the separate influences of the original sources. Importantly, the number of these original release sources is typically unknown, which hinders reliability of the classical inverse-model analyses. To address this challenge, we present here a novel hybrid method for identification of the unknown number of release sources. Our hybrid method, called HNMF, couples unsupervised learning based on Nonnegative Matrix Factorization (NMF) and inverse-analysis Green functions method. HNMF synergistically performs decomposition of the recorded mixtures, finds the number of the unknown sources and uses the Green function of advection-diffusion equation to identify their characteristics. In the paper, we introduce the method and demonstrate that it is capable of identifying the advection velocity and dispersivity of the medium as well as the unknown number, locations, and properties of various sets of synthetic release sources with different space and time dependencies, based only on the recorded data. HNMF can be applied directly to any problem controlled by a partial-differential parabolic equation where mixtures of an unknown number of sources are measured at multiple locations.
△ Less
Submitted 23 March, 2018; v1 submitted 12 December, 2016;
originally announced December 2016.
-
Hybridization of Higgs modes in a bond-density-wave state in cuprates
Authors:
Zachary M. Raines,
Valentin G. Stanev,
Victor M. Galitski
Abstract:
Recently, several groups have reported observations of collective modes of the charge order present in underdoped cuprates. Motivated by these experiments, we study theoretically the oscillations of the order parameters, both in the case of pure charge order, and for charge order coexisting with superconductivity. Using a hot-spot approximation we find in the coexistence regime two Higgs modes ari…
▽ More
Recently, several groups have reported observations of collective modes of the charge order present in underdoped cuprates. Motivated by these experiments, we study theoretically the oscillations of the order parameters, both in the case of pure charge order, and for charge order coexisting with superconductivity. Using a hot-spot approximation we find in the coexistence regime two Higgs modes arising from hybridization of the amplitude oscillations of the different order parameters. One of them has a minimum frequency that is within the single particle energy gap and which is a non-monotonic function of temperature. The other -- high-frequency -- mode is smoothly connected to the Higgs mode in the single-order-parameter region, but quickly becomes overdamped in the case of coexistence. We explore an unusual low-energy damping channel for the collective modes, which relies on the band reconstruction caused by the coexistence of the two orders. For completeness, we also consider the damping of the collective modes originating from the nodal quasiparticles. At the end we discuss some experimental consequences of our results.
△ Less
Submitted 13 August, 2015;
originally announced August 2015.
-
Anomalous proximity effects at the interface of s and s+- superconductors
Authors:
Valentin G. Stanev,
Alexei E. Koshelev
Abstract:
We study proximity effects close to a boundary between s and s+- superconductors. Frustration, caused by interaction of the s-wave gap parameter with the opposite-sign gaps of s+- superconductor, leads to several anomalous features. In the case of strong frustration a nontrivial time-reversal-symmetry breaking (TRSB) state, with nonzero phase angles between all gap parameters, is possible. In a mo…
▽ More
We study proximity effects close to a boundary between s and s+- superconductors. Frustration, caused by interaction of the s-wave gap parameter with the opposite-sign gaps of s+- superconductor, leads to several anomalous features. In the case of strong frustration a nontrivial time-reversal-symmetry breaking (TRSB) state, with nonzero phase angles between all gap parameters, is possible. In a more typical state, the s-wave order parameter is aligned with one of the s+- gaps. The other (anti-aligned) gap induces negative feature in the s-wave density of states, which can serve as a fingerprint of s+- state. Another consequence of the frustration is an extended region in the parameter space in which s-wave superconductivity is suppressed, despite being in contact with nominally stronger superconductor. This negative proximity effect is always present for the TRSB state, but extends even into the aligned states. We study these effects within a simple microscopic model assuming dirty limit in all bands, which allows us to model the system in terms of minimum number of the most relevant parameters. The described anomalous features provide a route to establishing the possible s+- state in the iron-based superconductors
△ Less
Submitted 27 October, 2012; v1 submitted 23 July, 2012;
originally announced July 2012.
