International Business Machines Corporation

Pascal Pompey

Michael Wurst

Olivier Verscheure

Mathieu Sinn

Francesco Fusco

Bei Chen

Eric P Bouillet

Thanh Lam Hoang

Rahul Nair

Christoph Lingenfelder

Randall Lee Cogill

John Sheehan

Marco Laumanns

Jakub Marecek

Dimitrios Mavroeidis

Carlos A Alzate Perez

Bradley Eck

William Karol Lynch

Vincent Lonij

Ulrike Fischer

Tim Nonner

Jacint Szabo

Karol W Lynch

Embodiments for ensemble policy generation for prediction systems by a processor. Policies are generated and/or derived for a set of ensemble models to predict a plurality of target variables for streaming data such that the plurality of policies enables dynamic adjustment of the prediction system. One or more of the policies are updated according to one or more error states of the set of ensemble models.

Ensemble model policy generation for prediction systems

Embodiments for self-managed adaptable models for prediction systems by one or more processors. One or more adaptive models may be applied to data streams from a plurality of data sources according to one or more data recipes such that the one or more adaptive models predict a plurality of target variables.

Self-managed adaptable models for prediction systems

Embodiments for automated data exploration and validation by a processor. One or more optimal data flows are provided in response to a query for one or more heterogeneous data sources according to an inference model based on a knowledge graph a plurality of data flows between one or more heterogeneous data sources relating to the query. An analytical flow is provided for one or more of the plurality of data flows for those of the one or more heterogeneous data sources that are undetected, and two or more of the one or more of the plurality of data flows are aggregated or disaggregated for the one or more heterogeneous data sources that are nested within the knowledge graph. One or more criteria is received from a user via an interactive graphical user interface (GUI) to use for defining the one or more optimal data flows.

Automated data exploration and validation

The cost of data-mining is estimated where data-mining services are delivered via a distributed computing system environment. System requirements are estimated for a particular data-mining task for an input data set having specified properties. Estimating system requirements includes applying a partial learning tool to operate on sample data from the input data set.

Estimating the cost of data-mining services

Techniques that facilitate semantic and time series analysis using machine learning are provided. In one example, a system includes a data analysis component, a prediction component and a learning component. The data analysis component that establishes one or more relationships between one or more elements of semantic data, including one or more time series identifiers, and one or more elements of time series data in a relationship database. The prediction component generates one or more advisory outputs, wherein generation of the one or more advisory outputs is performed in response to a trigger event. A learning component that determines the one or more relationships in the relationship database, wherein determination of the one or more relationships is based on information indicative of whether the advisory outputs satisfy a defined criterion.

Machine learning for time series using semantic and time series data

Techniques that facilitate semantic and time series analysis using machine learning are provided. In one example, a system includes a data analysis component, a prediction component and a learning component. The data analysis component that establishes one or more relationships between one or more elements of semantic data, including one or more time series identifiers, and one or more elements of time series data in a relationship database. The prediction component generates one or more advisory outputs, wherein generation of the one or more advisory outputs is performed in response to a trigger event, a learning component that determines the one or more relationships in the relationship database, wherein determination of the one or more relationships is based on information indicative of whether the advisory outputs satisfy a defined criterion.

Performing data clustering in a model property vector space. Input data is received comprising a plurality of data instances in a data vector space. A model property vector specification is defined for a model vector. Information is identified from the input data, and a model property vector is created in the model property vector space for each of the plurality of data instances. A target number of clusters is identified and used to perform a data clustering procedure. An output is generated comprising a plurality of data segments and one or more clustering rules. For each data cluster, a predictive model is constructed for each data segment of the plurality of data segments.

Predictive model clustering

Embodiments for automated data exploration and validation by a processor. One or more optimal data flows are provided in response to a query for one or more heterogeneous data sources according to an inference model based on a knowledge graph of heterogeneous data source relationships, a plurality of data flows between one or more heterogeneous data sources relating to the query, and an ontology of concepts and representing a domain knowledge of the one or more heterogeneous data sources.

In an approach for adaptive sampling of smart meter data, a computer retrieves one or more balancing constraints associated with one or more smart meter sensors. The computer retrieves meter sensor data from the one or more smart meter sensors according to the one or more balancing constraints. The computer determines a subsample of the meter sensor data based, at least in part, on one or more similar consumption patterns of meter sensor data, and then transmits the subsample of the meter sensor data to an optimization engine for use in solving an optimization problem.

Adaptive sampling of smart meter data

Embodiments for network reconstruction from message data by a processor. A digital map may be created using one or more messages of a plurality of vehicles obtained at a plurality of control points of a route network. The digital map may be analyzed to estimate a feasibility of simultaneous trajectories of the plurality of vehicles between selected locations in the route network.

Network map reconstruction from message data

In an approach for adaptive sampling of smart meter data, a computer retrieves one or more balancing constraints associated with one or more smart meter sensors. The computer retrieves meter sensor data from the one or more smart meter sensors according to the one or more balancing constraints. The computer determines a subsample of the meter sensor data, and then transmits the subsample of the meter sensor data to an optimization engine for use in solving an optimization problem.

Techniques for iterative feature extraction using domain knowledge are provided. In one aspect, a method for feature extraction is provided. The method includes the following steps. At least one query to predict at least one future value of a given value series based on a statistical model is received. At least two predictions of the future value are produced fulfilling at least the properties of 1) each being as probable as possible given the statistical model and 2) being mutually divert (in terms of numerical distance measure). A user is queried to select one of the predictions. The user may be queried for textual annotations for the predictions. The annotations may be used to identify additional covariates to create an extended set of covariates. The extended set of covariates may be used to improve the accuracy of the statistical model.

Growing community of inventors

Nanterre, France

Pascal Pompey