-
Compressed Sensing of Compton Profiles for Fermi Surface Reconstruction: Concept and Implementation
Authors:
J. Otsuki,
K. Yoshimi,
Y. Nakanishi-Ohno,
M. Sekania,
L. Chioncel,
M. Mizumaki
Abstract:
Compton scattering is a well-established technique that can provide detailed information about electronic states in solids. Making use of the principle of tomography, it is possible to determine the Fermi surface from sets of Compton-scattering data with different scattering axes. Practical applications, however, are limited due to long acquisition time required for measuring along enough number o…
▽ More
Compton scattering is a well-established technique that can provide detailed information about electronic states in solids. Making use of the principle of tomography, it is possible to determine the Fermi surface from sets of Compton-scattering data with different scattering axes. Practical applications, however, are limited due to long acquisition time required for measuring along enough number of scattering directions. Here, we propose to overcome this difficulty using compressed sensing. Taking advantage of a hidden sparsity in the momentum distribution, we are able to reconstruct the three-dimensional momentum distribution of bcc-Li, and identify the Fermi surface with as little as 14 directions of scattering data with unprecedented accuracy. This compressed-sensing approach will permit further wider applications of the Compton scattering experiments.
△ Less
Submitted 9 November, 2022; v1 submitted 14 October, 2022;
originally announced October 2022.
-
Statistical mechanical analysis of sparse linear regression as a variable selection problem
Authors:
Tomoyuki Obuchi,
Yoshinori Nakanishi-Ohno,
Masato Okada,
Yoshiyuki Kabashima
Abstract:
An algorithmic limit of compressed sensing or related variable-selection problems is analytically evaluated when a design matrix is given by an overcomplete random matrix. The replica method from statistical mechanics is employed to derive the result. The analysis is conducted through evaluation of the entropy, an exponential rate of the number of combinations of variables giving a specific value…
▽ More
An algorithmic limit of compressed sensing or related variable-selection problems is analytically evaluated when a design matrix is given by an overcomplete random matrix. The replica method from statistical mechanics is employed to derive the result. The analysis is conducted through evaluation of the entropy, an exponential rate of the number of combinations of variables giving a specific value of fit error to given data which is assumed to be generated from a linear process using the design matrix. This yields the typical achievable limit of the fit error when solving a representative $\ell_0$ problem and includes the presence of unfavourable phase transitions preventing local search algorithms from reaching the minimum-error configuration. The associated phase diagrams are presented. A noteworthy outcome of the phase diagrams is that there exists a wide parameter region where any phase transition is absent from the high temperature to the lowest temperature at which the minimum-error configuration or the ground state is reached. This implies that certain local search algorithms can find the ground state with moderate computational costs in that region. Another noteworthy result is the presence of the random first-order transition in the strong noise case. The theoretical evaluation of the entropy is confirmed by extensive numerical methods using the exchange Monte Carlo and the multi-histogram methods. Another numerical test based on a metaheuristic optimisation algorithm called simulated annealing is conducted, which well supports the theoretical predictions on the local search algorithms. In the successful region with no phase transition, the computational cost of the simulated annealing to reach the ground state is estimated as the third order polynomial of the model dimensionality.
△ Less
Submitted 10 September, 2018; v1 submitted 29 May, 2018;
originally announced May 2018.
-
Exhaustive search for sparse variable selection in linear regression
Authors:
Yasuhiko Igarashi,
Hikaru Takenaka,
Yoshinori Nakanishi-Ohno,
Makoto Uemura,
Shiro Ikeda,
Masato Okada
Abstract:
We propose a K-sparse exhaustive search (ES-K) method and a K-sparse approximate exhaustive search method (AES-K) for selecting variables in linear regression. With these methods, K-sparse combinations of variables are tested exhaustively assuming that the optimal combination of explanatory variables is K-sparse. By collecting the results of exhaustively computing ES-K, various approximate methods…
▽ More
We propose a K-sparse exhaustive search (ES-K) method and a K-sparse approximate exhaustive search method (AES-K) for selecting variables in linear regression. With these methods, K-sparse combinations of variables are tested exhaustively assuming that the optimal combination of explanatory variables is K-sparse. By collecting the results of exhaustively computing ES-K, various approximate methods for selecting sparse variables can be summarized as density of states. With this density of states, we can compare different methods for selecting sparse variables such as relaxation and sampling. For large problems where the combinatorial explosion of explanatory variables is crucial, the AES-K method enables density of states to be effectively reconstructed by using the replica-exchange Monte Carlo method and the multiple histogram method. Applying the ES-K and AES-K methods to type Ia supernova data, we confirmed the conventional understanding in astronomy when an appropriate K is given beforehand. However, we found the difficulty to determine K from the data. Using virtual measurement and analysis, we argue that this is caused by data shortage.
△ Less
Submitted 7 July, 2017;
originally announced July 2017.
-
Compressed sensing in scanning tunneling microscopy/spectroscopy for observation of quasi-particle interference
Authors:
Yoshinori Nakanishi-Ohno,
Masahiro Haze,
Yasuo Yoshida,
Koji Hukushima,
Yukio Hasegawa,
Masato Okada
Abstract:
We applied a method of compressed sensing to the observation of quasi-particle interference (QPI) by scanning tunneling microscopy/spectroscopy to improve efficiency and save measurement time. To solve an ill-posed problem owing to the scarcity of data, the compressed sensing utilizes the sparseness of QPI patterns in momentum space. We examined the performance of a sparsity-inducing algorithm cal…
▽ More
We applied a method of compressed sensing to the observation of quasi-particle interference (QPI) by scanning tunneling microscopy/spectroscopy to improve efficiency and save measurement time. To solve an ill-posed problem owing to the scarcity of data, the compressed sensing utilizes the sparseness of QPI patterns in momentum space. We examined the performance of a sparsity-inducing algorithm called least absolute shrinkage and selection operator (LASSO), and demonstrated that LASSO enables us to recover a double-circle QPI pattern of the Ag(111) surface from a dataset whose size is less than that necessary for the conventional Fourier transformation method. In addition, the smallest number of data required for the recovery is discussed on the basis of cross validation.
△ Less
Submitted 25 July, 2016;
originally announced July 2016.
-
Sparse approximation based on a random overcomplete basis
Authors:
Yoshinori Nakanishi-Ohno,
Tomoyuki Obuchi,
Masato Okada,
Yoshiyuki Kabashima
Abstract:
We discuss a strategy of sparse approximation that is based on the use of an overcomplete basis, and evaluate its performance when a random matrix is used as this basis. A small combination of basis vectors is chosen from a given overcomplete basis, according to a given compression rate, such that they compactly represent the target data with as small a distortion as possible. As a selection metho…
▽ More
We discuss a strategy of sparse approximation that is based on the use of an overcomplete basis, and evaluate its performance when a random matrix is used as this basis. A small combination of basis vectors is chosen from a given overcomplete basis, according to a given compression rate, such that they compactly represent the target data with as small a distortion as possible. As a selection method, we study the $\ell_0$- and $\ell_1$-based methods, which employ the exhaustive search and $\ell_1$-norm regularization techniques, respectively. The performance is assessed in terms of the trade-off relation between the representation distortion and the compression rate. First, we evaluate the performance analytically in the case that the methods are carried out ideally, using methods of statistical mechanics. Our result clarifies the fact that the $\ell_0$-based method greatly outperforms the $\ell_1$-based one. Second, we examine the practical performances of two well-known algorithms, orthogonal matching pursuit and approximate message passing, when they are used to execute the $\ell_0$- and $\ell_1$-based methods, respectively. Our examination shows that orthogonal matching pursuit achieves a much better performance than the exact execution of the $\ell_1$-based method, as well as approximate message passing. However, regarding the $\ell_0$-based method, there is still room to design more effective greedy algorithms than orthogonal matching pursuit. Finally, we evaluate the performances of the algorithms when they are applied to image data compression.
△ Less
Submitted 2 March, 2016; v1 submitted 7 October, 2015;
originally announced October 2015.
