Oracle America, Inc.

Oracle International Corporation

Sun Microsystems, Inc.

Oracle International Corporaiton

Kenny C Gross

Aleksey M Urmanov

Guang Chao Wang

Kalyanaraman Vaidyanathan

David K McElfresh

Anton A Bougaev

Ramakrishna C Dhanekula

Keith A Whisnant

Andrew J Lewis

Matthew T Gerdes

Edward R Wetherbee

Dieter Gawlick

Dan Vacar

Sanjeev Raghavendrachar Sondur

Zhen Hua Liu

Leoncio D Lopez

Alan Paul Wood

Steven F Zwinger

Mengying Li

Ashin George

Ayse K Coskun

Dustin R Garvey

Sajjit Thampy

Kenneth P Baclawski

Michael H S Dayringer

James Rohrkemper

Richard Paul Sonderegger

Timothy David Cline

Lik Wong

John J Cooley

Wendy Lu

David Brian Elting

Zexi Chen

Michael Dayringer

Rui Zhong

Sampanna Shahaji Salunke

Eric S Chan

Timothy Tsai

Prasanta Ranjan Dash

Andrew Brownsword

Jon D Greaves

Brandon Patrick Byers

Lawrence G Votta

Bradley R Williams

David Belanger

Anna Chystiakova

Mark Evan Mills

George Popescu

Hariharan Balasubramanian

Yujuan Bao

Benjamin P Franklin, Jr

Gregory A Cumberford

Cornell Thomas Eyford, Iii

Larry G Votta, Jr

Steven T Jeffreys

DeJun Li

Yixiu Liu

Michelle C Murrett

Robert Paul Masleid

Sungpack Hong

Ali Heydari

Stephen Ho

Chien Ouyang

Steven K Heller

Avishkar Misra

Georgios K Konstadinidis

Craig Schelp

Adel Ghoneimy

Kirk A Bradley

Andrew Vakhutinsky

Kishor S Trivedi

Guang-Tong Zhou

Saurabh Thapliyal

Becky X Li

Steve Lin

Glenn A Herdeg

David M Fishman

Shreya Singh

Neelesh Kumar Shukla

Roger E Blythe

Eugenio J Schuster

Filiz Gurtuna

Adam Porter

Kornelija Zgonc

Tahereh Masoumi

Peter L Wargo

Coleen L MacMillan

Xuemei Gao

Aakash K Chotrani

Yongqiang Zhang

Beiwen Guo

Debabrata Sarkar

Lawrence L Fumagalli, Jr

Yifan Wu

Guy G Michaeli

Mandip S Bhuller

Wei Jiang

William A Wimsatt

Stanley P Haag

Ashwini R More

Techniques for identifying a root cause of an operational result of a deterministic machine learning model are disclosed. A system applies a deterministic machine learning model to a set of data to generate an operational result, such as a prediction of a 'fault' or “no-fault” in the system. The set of data includes signals from multiple different data sources, such as sensors. The system applies an abductive model, generated based on the deterministic machine learning model, to the operational result. The abductive model identifies a particular set of data sources that is associated with the root cause of the operational result. The system generates a human-understandable explanation for the operational result based on the identified root cause.

Root cause analysis for deterministic machine learning model

Systems, methods, and other embodiments associated with frequency-domain resampling of time series are described. An example method includes generating a power spectrum for a first time series signal that is sampled inconsistently with a target sampling rate. Prominent frequencies are selected from the power spectrum. Sets of first phase factors that map the prominent frequencies to a frequency domain at first time points are generated. Coefficients are identified that relate the sets of first phase factors to values of the first time series signal at the first time points. Sets of second phase factors that map the prominent frequencies to a frequency domain at second time points are generated. A second time series signal that is resampled at the target sampling rate is generated by generating new values at the second time points from the coefficients and sets of second phase factors.

Frequency domain resampling of time series signals

Systems, methods, and other embodiments associated with eviction of weakly correlated signals from collections are described. In one embodiment, a mock signal that has random signal properties is generated. A mock correlation coefficient between the mock signal and a measured time series signal from a collection of measured time series signals is then generated. A discrimination value that indicates a weak signal correlation is then selected, based at least in part on the mock correlation coefficient. A first measured signal is then identified from the collection of measured time series signals that has the weak signal correlation by determining that a first correlation coefficient between the first measured signal and a second measured signal is weak based on the discrimination value. The first measured signal is then evicted from the collection of signals in response to the determination that the first measured signal has the weak signal correlation.

Eviction of weakly correlated signals from collections

Systems, methods, and other embodiments associated with acoustic detection of changes in mass of cargo carried by a vehicle are described herein. In one example method for acoustic cargo surveillance, a first acoustic output of a vehicle carrying cargo at a first time of surveillance of the vehicle is recorded. Then, a second acoustic output of the vehicle at a subsequent time in the surveillance of the vehicle carrying the cargo is recorded. A change in a mass of the cargo carried by the vehicle is acoustically detected based at least on an acoustic change between the first acoustic output and the second acoustic output. An electronic alert is generated that the mass of the cargo has changed based on the acoustic change.

Acoustic detection of cargo mass change

Systems, methods, and other embodiments associated with detecting feedback control instability in computer thermal controls are described herein. In one embodiment, a method includes executing a workload on the computing system, wherein the workload varies between a minimum and a maximum at a workload frequency. The method includes recording thermal telemetry from the computing system during execution of the workload. The method includes converting the recorded thermal telemetry into a frequency domain. The method includes detecting whether thermal control of the computing system exhibits feedback control instability based on dissimilarity in the frequency domain between the transformed thermal telemetry and the workload frequency. And, the method includes generating an electronic alert that indicates whether the thermal control of the computing device exhibits the feedback control instability.

Detection of feedback control instability in computing device thermal control

Systems, methods, and other embodiments associated with auditing the results of a machine learning model are described. In one embodiment, a method accesses original time series data and machine learning estimates of the original time series data. The method generates reconstituted time series data from the machine learning estimates by reversing operations of a machine learning model trained for generating the machine learning estimates from the original time series data. The method detects tampering (or corruption) in the original time series data based on a difference between the original time series data and reconstituted time series data. And, the method generates an electronic verification report that indicates whether the tampering (or corruption) is detected in the original time series data.

System and method for ensuring that the results of machine learning models can be audited

Systems, methods, and other embodiments associated with computer deepfake detection are described. In one embodiment, a method includes converting audio-visual content of a person delivering a speech into a set of time series signals. Residual time series signals of residuals that indicate an extent to which the time series signals differ from machine learning estimates of authentic delivery of the speech by the person are generated. Residual values from one synchronous observation of the residual time series signals are placed into an array of residual values for a point in time. A sequential analysis of the residual values of the array is performed to detect an anomaly in the residual values for the point in time. In response to detection of the anomaly, an alert that deepfake content is detected in the audio-visual content is generated.

Deepfake detection using synchronous observations of machine learning residuals

Systems, methods, and other embodiments associated with autonomous discrimination of operation vibration signals are described herein. In one embodiment, a method includes automatically choosing a plurality of vibration frequencies that vary in correlation with variation of a load on a monitored device. Vibration amplitudes for the plurality of vibration frequencies are monitored for incipient failure using a machine learning model. The machine learning model is trained to expect the vibration amplitudes to be consistent with undegraded operation of the monitored device. The incipient failure is detected where vibration amplitudes are not consistent with undegraded operation of the monitored device. An alert is then transmitted to suggest maintenance to prevent the incipient failure of the monitored device.

Autonomous discrimination of operation vibration signals

The disclosed embodiments provide a system that detects sensor anomalies in a univariate time-series signal. During a surveillance mode, the system receives the univariate time-series signal from a sensor in a monitored system. Next, the system performs a staggered-sampling operation on the univariate time-series signal to produce N sub-sampled time-series signals, wherein the staggered-sampling operation allocates consecutive samples from the univariate time-series signal to the N sub-sampled time-series signals in a round-robin ordering. The system then uses a trained inferential model to generate estimated values for the N sub-sampled time-series signals based on cross-correlations with other sub-sampled time-series signals. Next, the system performs an anomaly detection operation to detect incipient sensor anomalies in the univariate time-series signal based on differences between actual values and the estimated values for the N sub-sampled time-series signals. Whenever an incipient sensor anomaly is detected, the system generates a notification.

Staggered-sampling technique for detecting sensor anomalies in a dynamic univariate time-series signal

The disclosed system produces synthetic signals for testing machine-learning systems. During operation, the system generates a set of N composite sinusoidal signals, wherein each of the N composite sinusoidal signals is a combination of multiple constituent sinusoidal signals with different periodicities. Next, the system adds time-varying random noise values to each of the N composite sinusoidal signals, wherein a standard deviation of the time-varying random noise values varies over successive time periods. The system also multiplies each of the N composite sinusoidal signals by time-varying amplitude values, wherein the time-varying amplitude values vary over successive time periods. Finally, the system adds time-varying mean values to each of the N composite sinusoidal signals, wherein the time-varying mean values vary over successive time periods. The time-varying random noise values, amplitude values and mean values can be selected through a roll-of-the-die process from a library of values, which are learned from industry-specific signals.

Signal synthesizer data pump system

A model-based approach to determining an optimal configuration for a data center may use an environmental chamber to characterize the performance of various data center configurations at different combinations of temperature and altitude. Telemetry data may be recorded from different configurations as they execute a stress workload at each temperature/altitude combination, and the telemetry data may be used to train a corresponding library of models. When a new data center is being configured, the temperature/altitude of the new data center may be used to select a pre-trained model from a similar temperature/altitude. Performance of the current configuration can be compared to the performance of the model, and if the model performs better, a new configuration based on the model may be used as an optimal configuration for the data center.

Pattern-recognition enabled autonomous configuration optimization for data centers

Systems, methods, and other embodiments associated with acoustic fingerprint identification of devices are described. In one embodiment, a method includes generating a target acoustic fingerprint from acoustic output of a target device. A similarity metric is generated that quantifies similarity of the target acoustic fingerprint to a reference acoustic fingerprint of a reference device. The similarity metric is compared to a threshold. In response to a first comparison result of the comparing of the similarity metric to the threshold, the target device is indicated to match the reference device. In response to a second comparison result of the comparing of the similarity metric to the threshold, it is indicated that the target device does not match the reference device.

Acoustic fingerprinting

The disclosed embodiments provide a system that performs seasonality-compensated prognostic-surveillance operations for an asset. During operation, the system obtains time-series sensor signals gathered from sensors in the asset during operation of the asset. Next, the system identifies seasonality modes in the time-series sensor signals. The system then determines frequencies and phase angles for the identified seasonality modes. Next, the system uses the determined frequencies and phase angles to filter out the seasonality modes from the time-series sensor signals to produce seasonality-compensated time-series sensor signals. The system then applies an inferential model to the seasonality-compensated time-series sensor signals to detect incipient anomalies that arise during operation of the asset. Finally, when an incipient anomaly is detected, the system generates a notification regarding the anomaly.

Compensating for out-of-phase seasonality modes in time-series signals to facilitate prognostic-surveillance operations

We disclose a system that executes an inferential model in VRAM that is embedded in a set of graphics-processing units (GPUs). The system obtains execution parameters for the inferential model specifying: a number of signals, a number of training vectors, a number of observations and a desired data precision. It also obtains one or more formulae for computing memory usage for the inferential model based on the execution parameters. Next, the system uses the one or more formulae and the execution parameters to compute an estimated memory footprint for the inferential model. The system uses the estimated memory footprint to determine a required number of GPUs to execute the inferential model, and generates code for executing the inferential model in parallel while efficiently using available memory in the required number of GPUs. Finally, the system uses the generated code to execute the inferential model in the set of GPUs.

Determining memory requirements for large-scale ml applications to facilitate execution in GPU-embedded cloud containers

A double-blind comparison is performed between prognostic-surveillance systems, which are located on a local system and a remote system. During operation, the local system inserts random faults into a dataset to produce a locally seeded dataset, wherein the random faults are inserted into random signals at random times with variable fault signatures. Next, the local system exchanges the locally seeded dataset with a remote system, and in return receives a remotely seeded dataset, which was produced by the remote system by inserting different random faults into the same dataset. Next, the local system uses a local prognostic-surveillance system to analyze the remotely seeded dataset to produce locally detected faults. Finally, the local system determines a performance of the local prognostic-surveillance system by comparing the locally detected faults against actual faults in the remotely seeded fault information. The remote system similarly determines a performance of a remote prognostic-surveillance system.

Using a double-blind challenge to evaluate machine-learning-based prognostic-surveillance techniques

Growing community of inventors

San Diego, CA, United States of America

Kenny C Gross