公开(公告)号：WO2022079548A1

公开(公告)日：2022-04-21

申请号：PCT/IB2021/059161

申请日：2021-10-06

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： VALENTINE, Brian ,  EGGLESTON, John, Dale ,  BOISVERT, Brent ,  ALLEN, Michael ,  LUCKS, Michael ,  DREW, Brendon

IPC: G06F9/445

Abstract: A processor receives input data comprising: (i) a code level for an update, (ii) a scheduled time for the update; (iii) a target system for the update, and (iv) authorization data, wherein the authorization data: (i) allows for scheduling of the update and (ii) is provided via a channel external from a connection to the target system without a inbound connection. A processor receives a set of data from the target system. A processor, responsive to receiving the set of data from the target system, sends a response packet to the target system that includes the input data. A processor receives, at the scheduled time, a request to process the update. A processor, responsive to the request, sends code for processing the update corresponding to the code level for the update. A processor receives status messages corresponding to progress of the update.

公开(公告)号：WO2022079529A1

公开(公告)日：2022-04-21

申请号：PCT/IB2021/058735

申请日：2021-09-24

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： QUADER, Shaikh, Shahriar ,  JARAMILLO DURAN, Nicolas, Andres ,  MUKHOPADHYAY, Sumona ,  PAPANGELIS, Emmanouil ,  LITOIU, Marin ,  KALMUK, David ,  MIERZEJEWSKI, Peter

IPC: G06F16/90

Abstract: A DBMS training subsystem trains a DBMS workload-manager model with training data identifying resources used to execute previous DBMS data-access requests. The subsystem integrates each request's high-level features and compile-time operations into a vector and dusters similar vectors into templates. The requests are divided into workloads each represented by a training histogram that describes the distribution of templates associated with the workload and identifies the total amounts and types of resources consumed when executing the entire workload. The resulting knowledge is used to train the model to predict production resource requirements by: i) organizing production queries into candidate workloads; ii) deriving for each candidate a histogram similar in form and function to the training histograms; iii) using the newly derived histograms to predict each candidate's resource requirements; iv) selecting the candidate with the greatest resource requirements capable of being satisfied with available resources; and v) executing the selected workload.

公开(公告)号：WO2022034475A1

公开(公告)日：2022-02-17

申请号：PCT/IB2021/057325

申请日：2021-08-09

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： WANG, Dakuo ,  GAN, Chuang ,  BRAMBLE, Gregory ,  AMINI, Lisa ,  SAMULOWITZ, Horst Cornelius ,  KATE, Kiran ,  CHEN, Bei ,  WISTUBA, Martin ,  EVFIMIEVSKI, Alexandre ,  KATSIS, Ioannis ,  LI, Yunyao ,  MALOSSI, Adelmo Cristiano Innocenza ,  BARTEZZAGHI, Andrea ,  KAWAS, Ban ,  GURAJADA, Sairam ,  POPA, Lucian ,  PEDAPATI, Tejaswini ,  GRAY, Alexander

IPC: G06F16/2455

Abstract: A computer automatically selects a machine learning model pipeline using a meta-learning machine learning model. The computer receives ground truth data and pipeline preference metadata. The computer determines a group of pipelines appropriate for the ground truth data, and each of the pipelines includes an algorithm. The pipelines may include data preprocessing routines. The computer generates hyperparameter sets for the pipelines. The computer applies preprocessing routines to ground truth data to generate a group of preprocessed sets of said ground truth data and ranks hyperparameter set performance for each pipeline to establish a preferred set of hyperparameters for each of pipeline. The computer selects favored data features and applies each of the pipelines, with associated sets of preferred hyperparameters, to score the favored data features of the preprocessed ground truth data. The computer ranks pipeline performance and selects a candidate pipeline according to the ranking.

公开(公告)号：WO2022023861A1

公开(公告)日：2022-02-03

申请号：PCT/IB2021/056436

申请日：2021-07-16

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： KATSUKI, Takayuki

IPC: G06K9/00

Abstract: Techniques for fair anomaly detection include generating an anomaly detection model based on a Gaussian distribution of historical data, a mean vector of the Gaussian distribution, and a precision matrix of the Gaussian distribution. The mean vector and the precision matrix can be generated by reducing a function below a threshold, where the function can include the Gaussian distribution, a first regularization term configured to generate similar anomaly scores for inputs with similar fair features and independent of unfair features, and a second regularization term configured to generate similar anomaly localization scores for the inputs with the similar fair features and independent of the unfair features. The techniques further include inputting a new data to the anomaly detection model and generating an anomaly score and an anomaly localization score associated with the new data based on the Gaussian distribution, the mean vector, and the precision matrix.

公开(公告)号：WO2022013662A1

公开(公告)日：2022-01-20

申请号：PCT/IB2021/055808

申请日：2021-06-29

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： BUGGINS, Jack, Richard ,  TAHER, Luke ,  THAI, Vinh Tuan ,  MANNING, Ian

IPC: H04L29/06

Abstract: An approach for training a recurrent neural network to create a model for anomaly detection in the topology of a network is disclosed. The approach comprises, creating an embedding vector for each resource in the network based on applying an embedding algorithm to each resource of the network. A feature vector is then created for each change to a resource in the network based on one or more properties of the change. A recurrent neural network can thus be trained with the embedding vectors and the feature vectors to create a model for anomaly detection in the topology of the network.

公开(公告)号：WO2022009086A1

公开(公告)日：2022-01-13

申请号：PCT/IB2021/056039

申请日：2021-07-06

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： LEOBANDUNG, Effendi

IPC: H01L23/00

Abstract: Interposer-less multi-chip module are provided. In one aspect, an interposer-less multi-chip module includes: a substrate; a base film disposed on the substrate; and chips pressed into the base film, wherein top surfaces of the chips are coplanar. For instance, the chips can have varying thicknesses and are pressed into the base film to different depths such that top surfaces of the chips are coplanar. An interconnect layer having back-end-of line (BEOL) metal wiring can be present on the wafer over the chips. Methods of forming an interposer-less multi-chip module are also provided.

公开(公告)号：WO2022003469A1

公开(公告)日：2022-01-06

申请号：PCT/IB2021/055328

申请日：2021-06-16

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： CORVINELLI, Vincent ,  ALHAMID, Mohammed, Fahd ,  KALIL, Mohamad ,  ZUZARTE, Calisto

IPC: G06F17/00 ,  G06F16/24549 ,  G06F16/288 ,  G06N20/00

Abstract: In an approach to improve query optimization in a database management system, the methods identify opportunities for improvement in a cardinality estimate using a workload feedback process using a query feedback performed during query compilation. The methods identify correlations and relationships based on the structure of the query feedback and the runtime feedback performed, and collects data from the execution of a query to identify errors in estimates of the query optimizer. Further, the methods submit the query feedback and the runtime feedback to a machine learning engine to update a set of models. Additionally, the methods update a set of mode-s based on the submitted query feedback and runtime feedback, and output a new, updated, or re-trained model based on collected data from the execution of the query to identify the errors in estimates of the query optimizer, the submitted query feedback and the runtime feedback, or a trained generated mode.

公开(公告)号：WO2021260469A1

公开(公告)日：2021-12-30

申请号：PCT/IB2021/055065

申请日：2021-06-09

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： ZHANG, Da Wei ,  CHEN, Ke ,  SUN, Yu Ping ,  JIA, Hou Ping ,  MO, Xiaoguang

IPC: G10L13/00 ,  G10L13/02 ,  G10L13/08 ,  G06F3/012 ,  G10L13/033 ,  G10L13/047 ,  H04S7/303

Abstract: A computer converts each content sources from textual content to speech comprising a separate audio selection. The computer applies, to each audio selection, one or more speech attributes to specify the audio attributes that select a respective position of the respective audio selection from among multiple positions within a multidimensional sound space and audibly distinguish one or more characteristics of the respective audio selection from other audio selections, wherein the respective position of the respective audio selection reflects a rank of the respective audio selection as ordered by interest to a user. The computer outputs a simultaneous stream of the multiple audio selections to an audio output device for stereo play of the audio selections within the multiple positions within the multidimensional sound space to the user, with the multiple positions reflecting the content sources ordered by interest.

公开(公告)号：WO2021234522A1

公开(公告)日：2021-11-25

申请号：PCT/IB2021/054139

申请日：2021-05-14

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： DAS, Payel ,  CIPCIGAN, Flaviu ,  WADHAWAN, Kahini ,  PADHI, Inkit ,  VIJIL, Enara ,  CHEN, Pin-Yu ,  MOJSILOVIC, Aleksandra ,  SERCU, Tom ,  NOGUEIRA DOS SANTOS, Cicero

IPC: G16B20/30 ,  G16C20/30 ,  G16H20/13

Abstract: Techniques for filtering artificial intelligence (Al)-designed molecules for laboratory testing are provided. A computer implemented method can comprise selecting, by a system operatively coupled to a processor, a first subset of AI-designed molecules from a set of AI-designed molecules as candidate pharmaceutical agents based on classification of the Al-designed molecules using one or more classifiers. The method further comprises selecting, by the system, a second subset of the candidate pharmaceutical agents for wet laboratory testing based on evaluation of molecular interactions between the candidate pharmaceutical agents and one or more biological targets using one or more computer simulations.

公开(公告)号：WO2021229368A1

公开(公告)日：2021-11-18

申请号：PCT/IB2021/053781

申请日：2021-05-05

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION ,  IBM UNITED KINGDOM LIMITED ,  IBM (CHINA) INVESTMENT COMPANY LIMITED

Inventor： HATFIELD, Brian ,  GUPTA, Lokesh, Mohan ,  BORLICK, Matthew

IPC: G06F11/14

Abstract: In one method of write sort management, a write sort task related to write sorting a write list of data units to be destaged to storage, is assigned to a storage controller to improve the load balance among plural storage controllers. The available processing capacities of each of the storage controllers is determined by, for example, polling each of the storage controllers. A write sort task may then be assigned to a selected storage controller as a function of determined available processing capacities of each of the storage controllers to improve the load balance among the storage controllers.