# US 12,007,759 B2 · Geometric aging data reduction for machine learning applications > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,007,759 B2 - Document ID: US12007759B2 - Status: Granted - Kind: B2 - Application number: 17/361,189 - Filed: 2021-06-28 - Published: 2022-12-29 - Granted: 2024-06-11 - Earliest claimed: 2021-06-28 - Assignees: Oracle International Corporation - Applicants: Oracle International Corporation - Inventors: Dieter Gawlick; Matthew Torin Gerdes; Kirk Bradley; Anna Chystiakova; Zhen Hua Liu; Guang Chao Wang; Kenny C. Gross - Web page: https://patentgrep.com/patent/US12007759B2 ## CPC - G05B 23/024: {Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks} - G05B 23/0283: {Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] (preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06)} - G05B 23/0221: {Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods} - G06F 11/3006: {where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems (multiprogramming arrangements G06F9/46; allocation of resources G06F9/50)} - G06F 11/3409: {for performance assessment} - G06F 11/3452: {Performance evaluation by statistical analysis} - G06F 16/215: Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors - G06F 16/24568: {Data stream processing; Continuous queries} - G06F 18/15: Statistical pre-processing, e.g. techniques for normalisation or restoring missing data - G06F 18/213: Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods - G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade} - G06N 20/00: Machine learning - G16Y 40/20: Analytics; Diagnosis - G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G) - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G16Y: INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT] ## Abstract Techniques for geometric aging data reduction for machine learning applications are disclosed. In some embodiments, an artificial-intelligence powered system receives a first time-series dataset that tracks at least one metric value over time. The system then generates a second time-series dataset that includes a reduced version of a first portion of the time-series dataset and a non-reduced version of a second portion of the time-series dataset. The second portion of the time-series dataset may include metric values that are more recent than the first portion of the time-series dataset. The system further trains a machine learning model using the second time-series dataset that includes the reduced version of the first portion of the time-series dataset and the non-reduced version of the second portion of the time-series dataset. The trained model may be applied to reduced and/or non-reduced data to detect multivariate anomalies and/or provide other analytic insights. ## Claims ### Claim 1 A method comprising: receiving a first time-series dataset that tracks at least one metric value over time; generating a second time-series dataset that includes a reduced version of a first portion of the first time-series dataset and a non-reduced version of a second portion of the first time-series dataset, wherein the second portion of the first time-series dataset includes metric values that are more recent than the first portion of the first time-series dataset; determining one or more virtual metrics for the first portion of the first time-series dataset based on a non-reduced version of the first portion of the first time-series dataset, wherein the one or more virtual metrics are not discernable from the reduced version of the first portion of the first time-series dataset; and training at least one machine learning model using the second time-series dataset that includes the reduced version of the first portion of the first time-series dataset and the non-reduced version of the second portion of the first time-series dataset. ### Claim 2 (depends on 1) The method of claim 1, further comprising: generating the reduced version of the first portion of the first time-series dataset by merging adjacent values in the first portion of the first time-series dataset. ### Claim 3 (depends on 2) The method of claim 2, wherein merging adjacent values comprises averaging adjacent pairwise values. ### Claim 4 (depends on 2) The method of claim 2, further comprising: appending the second portion of the first time-series dataset to the first portion of the first time-series dataset. ### Claim 5 (depends on 1) The method of claim 1, wherein the second time-series dataset includes a version of a third portion of the first time-series dataset that has been reduced multiple times, wherein the third portion of the first time-series dataset includes metric values that are older than the first portion and second portion of the first time-series dataset. ### Claim 6 (depends on 1) The method of claim 1, further comprising: discarding the non-reduced version of the first portion of the dataset. ### Claim 7 (depends on 1) The method of claim 1, further comprising: generating the reduced version of the first portion of the first time-series dataset responsive to determining that a storage threshold has been reached. ### Claim 8 (depends on 1) The method of claim 1, further comprising: applying the machine learning model to at least a third time-series dataset; and generating an alert based on the output of the machine learning model indicative of anomalous behavior associated with the third time-series dataset. ### Claim 9 (depends on 1) The method of claim 1, wherein the one or more virtual metrics include at least one of a variance, mean, maximum, minimum, or standard deviation associated with the non-reduced version of the first portion of the first time-series dataset; wherein the one or more virtual metrics are preserved after discarding the non-reduced version of the first portion of the first time-series dataset and are used to train the at least one machine learning model. ### Claim 10 (depends on 1) The method of claim 1, wherein the one or more virtual metrics include at least one of a variance, mean, maximum, minimum, or standard deviation associated with the non-reduced version of the first portion of the first time-series dataset; wherein the one or more virtual metrics are preserved after discarding the non-reduced version of the first portion of the first time-series dataset and are used to train the at least one machine learning model. ### Claim 11 One or more non-transitory computer-readable media storing instructions, which, when executed by one or more hardware processors, cause: receiving a first time-series dataset that tracks at least one metric value over time; generating a second time-series dataset that includes a reduced version of a first portion of the first time-series dataset and a non-reduced version of a second portion of the first time-series dataset, wherein the second portion of the first time-series dataset includes metric values that are more recent than the first portion of the first time-series dataset; determining one or more virtual metrics for the first portion of the first time-series dataset based on a non-reduced version of the first portion of the first time-series dataset, wherein the one or more virtual metrics are not discernable from the reduced version of the first portion of the first time-series dataset; and training at least one machine learning model using the second time-series dataset that includes the reduced version of the first portion of the first time-series dataset and the non-reduced version of the second portion of the first time-series dataset. ### Claim 12 (depends on 11) The one or more non-transitory computer-readable media of claim 11, wherein the instructions further cause: generating the reduced version of the first portion of the first time-series dataset by merging adjacent values in the first portion of the first time-series dataset. ### Claim 13 (depends on 12) The one or more non-transitory computer-readable media of claim 12, wherein merging adjacent values comprises averaging adjacent pairwise values. ### Claim 14 (depends on 12) The one or more non-transitory computer-readable media of claim 12, wherein the instructions further cause: appending the second portion of the first time-series dataset to the first portion of the first time-series dataset. ### Claim 15 (depends on 11) The one or more non-transitory computer-readable media of claim 11, wherein the second time-series dataset includes a version of a third portion of the first time-series dataset that has been reduced multiple times, wherein the third portion of the first time-series dataset includes metric values that are older than the first portion and second portion of the first time-series dataset. ### Claim 16 (depends on 11) The one or more non-transitory computer-readable media of claim 11, wherein the instructions further cause: discarding the non-reduced version of the first portion of the dataset. ### Claim 17 (depends on 11) The one or more non-transitory computer-readable media of claim 11, wherein the instructions further cause: generating the reduced version of the first portion of the first time-series dataset responsive to determining that a storage threshold has been reached. ### Claim 18 (depends on 11) The one or more non-transitory computer-readable media of claim 11, wherein the instructions further cause: applying the machine learning model to at least a third time-series dataset; and generating an alert based on the output of the machine learning model indicative of anomalous behavior associated with the third time-series dataset. ### Claim 19 A system comprising: one or more hardware processors; one or more non-transitory computer-readable media storing instructions, which, when executed by one or more hardware processors, cause performance of operations comprising: receiving a first time-series dataset that tracks at least one metric value over time; generating a second time-series dataset that includes a reduced version of a first portion of the first time-series dataset and a non-reduced version of a second portion of the first time-series dataset, wherein the second portion of the first time-series dataset includes metric values that are more recent than the first portion of the first time-series dataset; determining one or more virtual metrics for the first portion of the first time-series dataset based on a non-reduced version of the first portion of the first time-series dataset, wherein the one or more virtual metrics are not discernable from the reduced version of the first portion of the first time-series dataset; and training at least one machine learning model using the second time-series dataset that includes the reduced version of the first portion of the first time-series dataset and the non-reduced version of the second portion of the first time-series dataset. ### Claim 20 (depends on 19) The system of claim 19, wherein the instructions further cause: generating the reduced version of the first portion of the first time-series dataset by merging adjacent values in the first portion of the first time-series dataset.