# US 11,113,611 B2 · Method and electronic apparatus for predicting electronic structure of material > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,113,611 B2 - Document ID: US11113611B2 - Status: Granted - Kind: B2 - Application number: 15/892,774 - Filed: 2018-02-09 - Published: 2019-02-28 - Granted: 2021-09-07 - Earliest claimed: 2017-08-28 - Assignees: Korea Institute of Science and Technology - Applicants: Korea Institute of Science and Technology - Inventors: Sang Soo Han; Byung Chul Yeo; Chansoo Kim; Donghun Kim - Web page: https://patentgrep.com/patent/US11113611B2 ## CPC - G06N 5/04: Inference or reasoning models - G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)} - G06F 17/16: Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)} - G06N 20/00: Machine learning - G16C 10/00: Computational theoretical chemistry, i.e. ICT specially adapted for theoretical aspects of quantum chemistry, molecular mechanics, molecular dynamics or the like - G16C 20/70: Machine learning, data mining or chemometrics - G16C 60/00: Computational materials science, i.e. ICT specially adapted for investigating the physical or chemical properties of materials or phenomena associated with their design, synthesis, processing, characterisation or utilisation - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G16C: COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE ## Abstract A method of predicting an electronic structure of a material by an electronic apparatus includes receiving user's input data about elements constituting the material; applying the received user's input data to a trained model for estimating a state density of the material; and outputting a first graph representing energy level-by-level state densities of the material output from the trained model, wherein the trained model is trained to generate the first graph based on a plurality of second graphs representing pre-calculated energy level-by-level state densities respectively corresponding to a plurality of pre-input data about elements of various materials and the plurality of pre-input data. ## Claims ### Claim 1 A method of predicting an electronic structure of a material by an electronic apparatus, the method comprising: receiving a user's input data about elements constituting the material; applying the received user's input data to a trained model for estimating a state density of the material; and outputting a first graph representing energy level-by-level state densities of the material output from the trained model, wherein the trained model is trained to generate the first graph based on a plurality of second graphs representing pre-calculated energy level-by-level state densities respectively corresponding to a plurality of pre-input data about elements of various materials and the plurality of pre-input data by converting the plurality of second graphs into recognition data, wherein the converting the plurality of second graphs comprises: converting the plurality of second graphs into a plurality of lattice images, respectively; converting the plurality of lattice images into a plurality of matrixes, respectively; calculating a covariance of the plurality of matrixes; calculating at least one eigenvector and at least one eigenvalue based on the calculated covariance; and determining at least one principal component representing a characteristic of the first graph based on the at least one eigenvector and the at least one eigenvalue; and wherein the trained model is trained to generate the first graph based on the at least one principal component. ### Claim 2 (depends on 1) The method of claim 1, wherein the trained model is used to determine a number of principal components used to generate the first graph based on the user's input data. ### Claim 3 (depends on 2) The method of claim 2, wherein the trained model is trained to determine the at least one principal component used to generate the first graph based on the determined number of principal components. ### Claim 4 (depends on 1) The method of claim 1, wherein the trained model is trained to acquire, from the received user's input data, a d-orbital electron number occupation rate of each of a plurality of elements constituting the material and determine a coefficient of the at least one principal component based on the acquired d-orbital electron number occupation rate. ### Claim 5 (depends on 4) The method of claim 4, wherein the d-orbital electron number occupation rate is determined based on a number of atoms and a d-orbital electron number of each of the elements constituting the material. ### Claim 6 An electronic apparatus for predicting an electronic structure of a material, the electronic apparatus comprising: a user input interface configured to receive a user's input data about elements constituting the material; a processor configured to apply the received user's input data to a trained model for estimating a state density of the material; and an output interface configured to output a first graph representing energy level-by-level state densities of the material output from the trained model, wherein the trained model is trained to generate the first graph based on a plurality of second graphs representing pre-calculated energy level-by-level state densities respectively corresponding to a plurality of pre-input data about elements of various materials and the plurality of pre-input data by converting the plurality of second graphs into recognition data, wherein the trained model is trained to: convert the plurality of second graphs into a plurality of lattice images, respectively, convert the plurality of lattice images into a plurality of matrixes, respectively, calculate a covariance of the plurality of matrixes, calculate at least one eigenvector and at least one eigenvalue based on the calculated covariance, determine at least one principal component representing a characteristic of the first graph based on the at least one eigenvector and the at least one eigenvalue and generate the first graph based on the at least one principal component. ### Claim 7 (depends on 6) The electronic apparatus of claim 6, wherein the trained model is trained to determine a number of principal components used to generate the first graph based on the user's input data. ### Claim 8 (depends on 7) The electronic apparatus of claim 7, wherein the trained model is trained to determine the at least one principal component used to generate the first graph based on the determined number of principal components. ### Claim 9 (depends on 6) The electronic apparatus of claim 6, wherein the trained model is trained to acquire, from the received user's input data, a d-orbital electron number occupation rate of each of a plurality of elements constituting the material and determine a coefficient of the at least one principal component based on the acquired d-orbital electron number occupation rate. ### Claim 10 (depends on 9) The electronic apparatus of claim 9, wherein the d-orbital electron number occupation rate is determined based on a number of atoms and a d-orbital electron number of each of constituent elements of the material. ### Claim 11 (depends on 1) A non-transitory computer-readable recording medium that stores a program that, when executed by a computer, performs the method of claim 1.