International Business Machines Corporation

Globalfoundries Inc.

Hewlett-packard Company

Arun Kwangil Iyengar

Paul Michael Dantzig

Isabelle Marie Rouvellou

James R H Challenger

Gong Su

Francis Nicholas Parr

Peter Hans Westerink

Norman Bobroff

Daniel Manuel Dias

George Prentice Copeland

Louis Degenaro

Thomas Arthur Mikalsen

Mustafa Canim

Richard Pervin King

Jian Yin

Sastry Duri

Ignacio Silva-Lepe

James Robert Harold Challenger

Liana Liyow Fong

Robert O Dryfoos

Dhavalkumar C Patel

Joshua J Milthorpe

Robert G Farrell

Erich Miles Nahum

John A Gunnels

Rajesh S Agarwalla

Karen A Witting

Mudhakar Srivatsa

Charles P Wright

Nianjun Zhou

Asser Nasreldin Tantawi

Michael H Dawson

Rahul P Akolkar

Daniela Rosu

Mark Holmes Linehan

Revathi Subramanian

Subbarao K Meduri

Wolfgang Segmuller

Shrey Shrivastava

Hongbo Jiang

Gerald A Spivak

Eric Michel Levy-Abegnoli

Ashish Kundu

Wei Tan

Michael Andrew Blake

Seetharami R Seelam

Qi Zhang

Keun Soo Yim

Shicong Meng

Robert Filepp

Matthew McClain

John Lewars

Judah M Diament

Junehwa Song

Ryan A Musa

Cristina Cornelio

Mariano Rodriguez Muro

Jeffrey Owen Kephart

Donna N Dillenberger

Hidayatullah Habeebullah Shaikh

Asit Dan

Anuradha Bhamidipaty

Emmanuel Yashchin

Graeme Johnson

Dung Tien Phan

Michael Haden Conner

Mark Richard Gambino

Chandrasekhara K Reddy

Paul Reed

Juan Du

Daniel Manual Dias

Manoj N Kumar

Frederick Douglis

Murthy V Devarakonda

Paul J Callaway

Renu Tewari

Achille Belly Fokoue-Nkoutche

Jia Zou

Malgorzata Steinder

Zehra Noman Sura

Vesselina K Papazova

Ian N Whalley

Nagui Halim

Rajat Mukherjee

Kenneth Douglas Klapproth

Yu Fen Yuan

Lei Gao

Christopher Shane Claussen

Agnes Shanthi Thiruthuvadoss

German Sergio Goldszmidt

Jason D Kohl

David Lars Ehnebuske

Guerney Douglass Holloway Hunt

James R Challenger

Claris Castillo

Barbara Jane Alspach McKee

Tim Bronson

Sunil Kiran Cutinho

Stephen M Goldman

Techniques for model evaluation and selection are provided. A plurality of models trained to generate predictions at each of a plurality of intervals is received, and a plurality of model ensembles, each specifying one or more of the plurality of models for each of the plurality of intervals, is generated. A test data set is received, where the test data set includes values for at least a first interval of the plurality of intervals and does not include values for at least a second interval of the plurality of intervals. A first model ensemble, of the plurality of model ensembles, is selected based on processing the test data set using each of the plurality of model ensembles.

Determining and selecting prediction models over multiple points in time using test data

Methods and systems to provide a form of probabilistic labeling to associate an outage with a disturbance, which could itself be either known based on the available data or unknown. In the latter case, labeling is especially challenging, as it necessitates the discovery of the disturbance. One approach incorporates a statistical change-point analysis to time-series events that correspond to service tickets in the relevant geographic sub-regions. The method is calibrated to separate the regular periods from the environmental disturbance periods, under the assumption that disturbances significantly increase the rate of loss-causing events. To obtain the probability that a given loss-causing event is related to an environmental disturbance, the method leverages the difference between the rate of events expected in the absence of any disturbances (baseline) and the rate of actually observed events. In the analysis, the local disturbances are identified and estimators of their duration and magnitude are provided.

Associating disturbance events to accidents or tickets

Disclosed embodiments provide techniques for estimating imputation algorithm performance. Multiple imputer algorithms are selected, and an evaluation of how well each of the imputer algorithms can estimate the missing data is performed. Disclosed embodiments obtain an imputer candidate dataset (ICD). The imputer candidate dataset is compared to the incomplete data range, and a similarity metric is determined between the data range and the ICD. When the similarity metric exceeds a predetermined threshold, an imputer evaluation dataset (IED) is created from the ICD by removing one or more data points from the ICD. Each imputer algorithm is evaluated by applying the IED to it, and computing an imputer evaluation metric based on its performance. The multiple imputer algorithms are ranked based on the imputer evaluation metric. The best ranked imputer algorithm can then be selected for use on the incomplete data range within the measurement dataset.

Data gap mitigation

Techniques for generating model ensembles are provided. A plurality of models trained to generate predictions at each of a plurality of intervals is received. A respective prediction accuracy of each respective model of the plurality of models is determined for a first interval of the plurality of intervals by processing labeled evaluation data using the respective model. Additionally, a model ensemble specifying one or more of the plurality of models for each of the plurality of intervals is generated, comprising selecting, for the first interval, a first model of the plurality of models based on (i) the respective prediction accuracies and (ii) at least one non-error metric.

Determining and selecting prediction models over multiple points in time

A processing system server and methods for performing asynchronous data store operations. The server includes a processor which maintains a cache of objects in memory of the server. The processor executes an asynchronous computation to determine the value of an object. In response to receiving a request for the object occurring before the asynchronous computation has determined the value of the object, a value of the object is returned from the cache. In response to receiving a request for the object occurring after the asynchronous computation has determined the value of the object, a value of the object determined by the asynchronous computation is returned. The asynchronous computation may comprise at least one future, such as a ListenableFuture, or a process or thread. The asynchronous computation may determine the value of the object by querying at least one additional server.

Providing data values using asynchronous operations and based on timing of occurrence of requests for the data values

Automated techniques are disclosed for minimizing communication between nodes in a system comprising multiple nodes for executing requests in which a request type is associated with a particular node. For example, a technique comprises the following steps. Information is maintained about frequencies of compound requests received and individual requests comprising the compound requests. For a plurality of request types which frequently occur in a compound request, the plurality of request types is associated to a same node. As another example, a technique for minimizing communication between nodes, in a system comprising multiple nodes for executing a plurality of applications, comprises the steps of maintaining information about an amount of communication between said applications, and using said information to place said applications on said nodes to minimize communication among said nodes.

Managing multileg transactions in distributed and parallel environments

An information processing system, computer readable storage medium, and a method for managing computer memory. The method includes deploying at least one memory allocator on a computing system which includes a multi-user computing system or a cloud computing system; dynamically allocating, by the at least one memory allocator, an amount of dynamically allocated memory for use by an executing program at runtime; and determining a cost to charge a user based on the amount of memory that has been allocated. Multiple billing methods can be used including: restricting the amount of dynamically allocated memory available to the user based on how much the user is paying; charging the user a fee which grows with an amount of dynamically allocated memory consumed; charging the user higher fees during higher periods of dynamically allocated memory usage; and charging the user a low fee for maintaining dynamically allocated memory usage below a threshold.

Charging users for computer memory usage

Improved techniques are disclosed for processing transactions or requests in a data processing system. For example, a method for processing requests in a system including a plurality of nodes includes the following steps. At least two nodes of the plurality of nodes receive a plurality of requests. The two nodes exchange information to determine an order for processing requests. The two nodes process the requests in accordance with the order. The order may include a total order or a partial order.

Highly ordered transaction processing

A method for transparently moving a block of memory with respect to an application using the block of memory, includes inserting, by a compiler, in an application that includes a memory allocation call, instructions for transparently moving a block of memory with respect to an application using the block of memory. The instructions include obtaining a first pointer returned by a memory allocator, where the first pointer points to an internal data structure, the internal data structure includes a read-write lock and a second pointer, and the second pointer points to an actual memory block. The instructions further include acquiring a read lock on a read-write lock in the internal data structure, before the first pointer is used by the application, obtaining the second pointer to the actual memory block, and dereferencing the second pointer to access the actual memory block for the application data.

Reducing fragmentation of computer memory

An adaptive monitoring method, system, and computer program product including an intelligent monitoring system which obtains at least one preference from a user, where the at least one preference includes at least one maximum monitoring overhead.

Adaptive monitoring of computing systems

A method, system, and computer program product for data compression in storage clients. In some embodiments, a storage client for accessing a storage service from a computer program is provided. A compression method is provided in the storage client to reduce a size of data objects. A frequency of compressing data from the computer program or modifying a compression algorithm based on assessing costs and benefits of compressing the data is varied.

Adaptive compression for data services

An information processing system, a computer readable storage medium, and a computer-implemented method, collect tables from a corpus of documents, convert the collected tables to flattened table format and organized to be searchable by schema-less queries. A method collects tables, extracts feature values from collected table data and collected table meta-data for each collected table. A table classifier classifies each collected table as being a type of table. Based on the classifying, the collected table is converted to a flattened table including table values that are the table data and the table meta-data of the collected table. Dependencies of the data values are mapped. The flattened table and mapped dependencies are stored in a triple store searchable by schema-less queries. The table classifier learns and improves its accuracy and reliability. Dependency information is maintained among a plurality of database tables. The dependency information can be updated at variable update frequency.

Automatic transformation of complex tables in documents into computer understandable structured format and managing dependencies

A processing system, computer program product, and methods for maintaining consistent values of objects in two or more stores. A method includes a data store associating an updated-version-request-time for each object of a plurality of objects wherein each object of the plurality of objects resides on the data store. The data store maintains a data structure ordering the plurality of objects by updated-version-request-times. The data store periodically examines the data structure to read updated-version-request-times. The data store makes a decision to notify another data store of a change to an object belonging to the plurality of objects based on an updated-version-request-time of the object. The data structure can be a balanced tree.

Selectively sending updated data values

Embodiments of the invention are directed a computer-implemented method for efficiently assessing data quality metrics. A non-limiting example of the computer-implemented method includes receiving, using a processor, a plurality of updates to data points in a data stream. The processor is further used to provide a plurality of data quality metrics (DQMs), and to maintain information on how much the plurality of DQMs are changing over time. The processor also maintains information on computational overhead for the plurality of DQMs, and also updates data quality information based on the maintained information.

Efficient real-time data quality analysis

A processing system, computer program product, and methods for maintaining consistent values of objects in two or more stores. A method includes a data store associating an updated-version-request-time for each object of a plurality of objects, where each object of the plurality of objects resides on a respective another data store. The data store maintains a data structure ordering the plurality of objects by updated-version-request-times. The data store periodically examines the data structure to read updated-version-request-times. The data store makes a decision to request an updated version of an object belonging to the plurality of objects based on an updated-version-request-time of the object. The data structure can be a balanced tree.

Selectively requesting updated data values

Growing community of inventors

Yorktown Heights, NY, United States of America

Arun Kwangil Iyengar