メーリングリストの皆様
東京大学生産技術研究所の羽田野と申します。
羽田野研究室では、下記の通りセミナーを開催致します。
皆様の奮ってのご参加をお待ちしております。
なお、当研究室におけるセミナー情報は、次のリンクよりご覧頂けます。
hatano-lab.iis.u-tokyo.ac.jp/seminar-j.html (English: hatano-lab.iis.u-tokyo.ac.jp/seminar-e.html)
＊本案内は複数のメーリングリストにお送りしております。重複して受け取られました方は何卒ご容赦下さい。
記
日時：２０１９年１２月１１日（水）１３時３０分〜 Wed, Nov 27, 1:30pm
場所：東京大学 生産技術研究所 千葉実験所 研究実験棟Ｉn210号室
道程：http://hatano-lab.iis.u-tokyo.ac.jp/access-j.html (English: hatano-lab.iis.u-tokyo.ac.jp/access-e.html)
講師：本山裕一 さん (東大物性研) Dr. Yuichi Motoyama (ISSP, U. Tokyo)
演題：Analytic continuation methods from imaginary time correlation function to real frequency spectrum
要旨：It is important to calculate dynamical quantities such as the spectrum function (density of states) and the dynamical structure factor by numerical methods because they reflect on the structure of low-energy excited states and they are observed by experiments, e.g., ARPES. By using the path-integral Monte Carlo method, these quantities can be calculated through the so-called “analytic continuation” (AC) method from the imaginary-time correlation functions such as the temperature Green’s function. The AC method, unfortunately, is well-known as ill-posed problem; the noise (the statistical error) of the correlation function (input) is exponentially magnified in the spectrum function (output). In order to overcome this instability, several AC methods have been developed in decades. For example, the SpM method adopts the sparse modeling method for reducing the noise, so that SpM gets a strong robustness. The SpM method, unfortunately, suffers from an unphysical oscill
ation due to the truncation. In this talk, I will introduce our new method, the SpM-Pade method, which is based on two existing methods, the Pade method and the SpM method. I will also review other AC methods, the maximum entropy method and the deep learning method.
———————————————-
羽田野直道
277-8574 千葉県柏市柏の葉5-1-5
東京大学生産技術研究所
Phone: 04-7136-6977
Internal: 58974, 66977
Fax: 04-7136-6978
hatano-lab.iis.u-tokyo.ac.jp/hatano/index-j.html
————————————————-
Computational Material Physics Mailing List
home: www.issp.u-tokyo.ac.jp/public/cmp/
archive: cmp-ml.issp.u-tokyo.ac.jp
twitter: https://twitter.com/cmp_ml
————————————————-