# US 12,001,440 B2 · Prediction device, prediction method, and prediction program > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,001,440 B2 - Document ID: US12001440B2 - Status: Granted - Kind: B2 - Application number: 17/621,228 - Filed: 2019-06-26 - Published: 2022-11-03 - Granted: 2024-06-04 - Earliest claimed: 2019-06-26 - Assignees: Nippon Telegraph and Telephone Corporation - Applicants: Nippon Telegraph and Telephone Corporation - Inventors: Daisuke Sato; Tatsushi Matsubayashi; Hiroyuki Toda - Web page: https://patentgrep.com/patent/US12001440B2 ## CPC - G06F 16/2477: {Temporal data queries} - G06F 16/2465: {Query processing support for facilitating data mining operations in structured databases} - G06N 20/00: Machine learning - G06Q 10/04: Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR ## Abstract A prediction device includes: a normal prediction unit configured to predict first prediction values for target ranges of respective models, based on respective pieces of spatio-temporal observation data including observation values for respective observation locations and respective observation times observed for a given time period and the respective models for predicting variations of the pieces of spatio-temporal observation data for the respective target ranges defined for at least one of an observation location or a period; a residual extraction unit configured to extract, based on residuals between the respective pieces of spatio-temporal observation data being observed and the models corresponding to the respective pieces of spatio-temporal observation data based on the pieces of spatio-temporal observation data and the models, the respective pieces of spatio-temporal observation data with residuals greater than or equal to a threshold value, as spatio-temporal residual data; a residual prediction unit configured to predict, based on the spatio-temporal residual data, second prediction values for the target ranges corresponding to the spatio-temporal residual data in accordance with a predetermined prediction technique; and a synthesis unit configured to output prediction values obtained by synthesizing the first prediction values and the second prediction values. ## Claims ### Claim 1 A prediction device comprising circuit configured to execute operations comprising: predicting first prediction values for target ranges of respective models, based on respective pieces of spatio-temporal observation data including observation values for respective observation locations and respective observation times observed for a given time period and the respective models for predicting variations of the pieces of spatio-temporal observation data for the respective target ranges defined for at least one of an observation location or a period; extracting, based on residuals between the respective pieces of spatio-temporal observation data being observed and the models corresponding to the respective pieces of spatio-temporal observation data based on the pieces of spatio-temporal observation data and the models, the respective pieces of spatio-temporal observation data with residuals based on a threshold value, as spatio-temporal residual data; predicting, based on the spatio-temporal residual data, second prediction values for the target ranges corresponding to the spatio-temporal residual data in accordance with a predetermined prediction technique; synthesizing the first prediction values and the second prediction values to generate output prediction values; and outputting the output prediction values. ### Claim 2 (depends on 1) The prediction device according to claim 1, the circuit further configured to execute operations comprising: registering, in a storage, flags for updating the respective models with the residuals equal to or greater than the threshold value. ### Claim 3 (depends on 2) The prediction device according to claim 2, the circuit further configured to execute operations comprising: referring to the respective flags registered in the storage; analyzing for the models corresponding to the flags, variations in the target ranges of the models by using respective observation values corresponding to the target ranges of the models stored in the storage; and updating the models. ### Claim 4 (depends on 3) The prediction device according to claim 3, wherein the analyzing variations in the target ranges of the models is based on one of regression including an autoregressive model or logistic regression. ### Claim 5 (depends on 1) The prediction device according to claim 1, wherein the models include a periodic model for one of: each day of the week, time of day. ### Claim 6 (depends on 1) The prediction device according to claim 1, wherein the spatio-temporal residual data excludes the respective pieces of spatio-temporal observation data with residuals less than the threshold value. ### Claim 7 (depends on 1) The prediction device according to claim 1, wherein the predetermined prediction technique includes one of a vector autoregressive model, reduced rank regression, or Gaussian process regression. ### Claim 8 A computer-implemented method for predicting data comprising: predicting first prediction values for target ranges of respective models, based on respective pieces of spatio-temporal observation data including observation values for respective observation locations and respective observation times observed for a given time period and the respective models for predicting variations of the pieces of spatio-temporal observation data for the respective target ranges defined for at least one of an observation location or a period; extracting, based on residuals between the respective pieces of spatio-temporal observation data being observed and the models corresponding to the respective pieces of spatio-temporal observation data based on the pieces of spatio-temporal observation data and the models, the respective pieces of spatio-temporal observation data with residuals based on a threshold value, as spatio-temporal residual data; predicting, based on the spatio-temporal residual data, second prediction values for the target ranges corresponding to the spatio-temporal residual data in accordance with a predetermined prediction technique; synthesizing the first prediction values and the second prediction values to generate prediction values; and outputting the prediction values. ### Claim 9 (depends on 8) The prediction method according to claim 8, further comprising: registering, in a storage, flags for updating the respective models with the residuals equal to or greater than the threshold value. ### Claim 10 (depends on 9) The prediction method according to claim 9, further comprising: referring to the respective flags registered in the storage; and analyzing, for the models corresponding to the flags, variations in the target ranges of the models by using respective observation values corresponding to the target ranges of the models stored in the storage; and updating the models. ### Claim 11 (depends on 10) The computer-implemented method according to claim 10, wherein the analyzing variations in the target ranges of the models is based on one of regression including an autoregressive model or logistic regression. ### Claim 12 (depends on 8) The computer-implemented method according to claim 8, wherein the models include a periodic model for one of: each day of the week, time of day. ### Claim 13 (depends on 8) The computer-implemented method according to claim 8, wherein the spatio-temporal residual data excludes the respective pieces of spatio-temporal observation data with residuals less than the threshold value. ### Claim 14 (depends on 8) The computer-implemented method according to claim 8, wherein the predetermined prediction technique includes one of a vector autoregressive model, reduced rank regression, or Gaussian process regression. ### Claim 15 A computer-readable non-transitory recording medium storing computer-executable program instructions that when executed by a processor cause a computer system to execute operations comprising: predicting first prediction values for target ranges of respective models, based on respective pieces of spatio-temporal observation data including observation values for respective observation locations and respective observation times observed for a given time period and the respective models for predicting variations of the pieces of spatio-temporal observation data for the respective target ranges defined for at least one of an observation location or a period; extracting, based on residuals between the respective pieces of spatio-temporal observation data being observed and the models corresponding to the respective pieces of spatio-temporal observation data based on the pieces of spatio-temporal observation data and the models, the respective pieces of spatio-temporal observation data with residuals based on a threshold value, as spatio-temporal residual data; predicting, based on the spatio-temporal residual data, second prediction values for the target ranges corresponding to the spatio-temporal residual data in accordance with a predetermined prediction technique; synthesizing the first prediction values and the second prediction values to generate prediction values; and outputting the prediction values. ### Claim 16 (depends on 7) The computer-readable non-transitory recording medium according to claim 7, the computer-executable program instructions when executed further causing the system to execute operations comprising: registering, in a storage, flags for updating the respective models with the residuals equal to or greater than the threshold value. ### Claim 17 (depends on 15) The computer-readable non-transitory recording medium according to claim 15, the computer-executable program instructions when executed further causing the system to execute operations comprising: referring to the respective flags registered in the storage; analyzing, for the models corresponding to the flags, variations in the target ranges of the models by using respective observation values corresponding to the target ranges of the models stored in the storage; and updating the models. ### Claim 18 (depends on 15) The computer-readable non-transitory recording medium according to claim 15, wherein the models include a periodic model for one of: each day of the week, time of day. ### Claim 19 (depends on 15) The computer-readable non-transitory recording medium according to claim 15, wherein the spatio-temporal residual data excludes the respective pieces of spatio-temporal observation data with residuals less than the threshold value. ### Claim 20 (depends on 15) The computer-readable non-transitory recording medium according to claim 15, wherein the predetermined prediction technique includes one of a vector autoregressive model, reduced rank regression, or Gaussian process regression.