Monsanto Technology LLC

Intuit, Inc.

Cisco Technology, Inc.

Sambarta Dasgupta

Srinivas Phani Kumar Chavali

Zi Wang

Colin R Dillard

Nalini Polavarapu

Jean-Philippe Vasseur

Shashank Shashikant Rao

Vinay Kumar Kolar

Anthony Paul Kovacs

Grégory Mermoud

Jorge Moran

Sricharan Kallur Palli Kumar

Yiduo Zhan

Silvano Assanga Ocheya

Aviral Shukla

Sricharan Kumar

Qianni Dong

Benjamin Bruce Stewart-Brown

Chong Yu

Mahdi Jadaliha

Jesse B Grote

Anand Pramod Deshmukh

Hui Zhang

Humberto Ignacio Gutierrez Gaitan

Xiaobo Zhou

Sricharan Kallur Kumar

Ashok N Srivastava

Mukund Yelahanka Raghuprasad

Ji Chen

Sameeksha Khillan

Debasish Das

Hongwei Luo

Colin Dillard

Ashok Srivastava

Jorge Luis Moran

Exemplary methods for use in identifying crosses for use in plant breeding are disclosed. One exemplary method includes generating population prediction scores for each potential cross within a set of potential crosses, where each population prediction score is associated with a prediction of commercial success for the associated potential cross within the set of potential crosses. The method also includes selecting a subgroup of potential crosses, based on thresholds associated with the population prediction scores for the set of potential crosses. The exemplary method further includes selecting multiple target crosses from the subgroup of potential crosses based on a genetic relatedness of the parents in the subgroup of potential crosses, and directing ones of the selected target crosses into a breeding pipeline, thereby providing crosses to the breeding pipeline based, at least in part, on commercial success of parents included in the selected ones of the filtered crosses.

Methods for identifying crosses for use in plant breeding

Exemplary systems and methods are disclosed for allocating resources in a breeding pipeline to multiple origins. One exemplary method includes accessing a data structure including data representative of multiple origins, in which the data includes, for each of the multiple origins, a trait performance expression or genotypic component information. The exemplary method further includes determining a resource allocation, which allocates n resources among the multiple origins based on a probability associated with the trait performance expressions and/or the genotypic components for the origins, and then allocating the n resources in the breeding pipeline for the multiple origins, based on the determined resource allocation.

Methods and systems for use in implementing resources in plant breeding

A routing processor implements a multi-stage prescriptive routing model engine based on harvest input data relating to the harvesting of crops at a plurality of locations by a plurality of combines and based on a harvest characteristic representing an attribute of the crops to be harvested by the combines. The multi-stage prescriptive routing model engine generates a combine routing program prescribing the movement of each combine between the locations and includes a demand stage configured to identify combine harvesting demand as a function of the harvest input data and the harvest characteristic and a scheduling stage configured to generate the combine routing program as a function of the harvesting demand.

Systems and methods for combine routing

In one embodiment, a device identifies a timeseries motif present in a plurality of timeseries of performance metrics for a plurality of paths in a network. The device retrieves, based on the timeseries motif, device-level telemetry data from networking devices along the plurality of paths. The device determines a root cause of the timeseries motif by correlating the timeseries motif with the device-level telemetry data. The device provides an indication of the timeseries motif and its root cause for display by a user interface.

Network-based mining approach to root cause impactful timeseries motifs

Certain aspects of the present disclosure provide techniques for performing finite rank deep kernel learning. In one example, a method for performing finite rank deep kernel learning includes receiving a training dataset; forming a set of embeddings by subjecting the training dataset to a deep neural network; forming, from the set of embeddings, a plurality of dot kernels; linearly combining the plurality of dot kernels to form a composite kernel for a Gaussian process; receiving live data from an application; and predicting a plurality of values and a plurality of uncertainties associated with the plurality of values simultaneously using the composite kernel.

Finite rank deep kernel learning with linear computational complexity

In one embodiment, a device computes time series dynamics for a performance metric of a path in a network used to convey traffic for an online application. The device matches those time series dynamics to one or more dynamics categories. The device makes a determination as to whether the path in the network is anomalous, based on the one or more dynamics categories. The device provides, based on the determination, an indication that the path in the network is anomalous for display.

Automatic detection and tracking of anomalous rectifiable paths using time-series dynamics

In one embodiment, a device generates a plurality of smoothed timeseries by applying smoothing envelopes of different durations to a timeseries of a path metric for a path in a network that is used to convey traffic of an online application. The device uses the plurality of smoothed timeseries and the timeseries of the path metric to make predictions as to whether the path will provide an unacceptable user experience in the online application. The device selects a smoothing envelope of a particular duration, by comparing performance metrics for the predictions. The device uses a timeseries of the path metric smoothed using the smoothing envelope of the particular duration to make predictive routing decisions in the network for the traffic of the online application.

Predictive network routing with dynamic smoothing envelope creation for noisy network timeseries

Systems and methods for forecasting cashflows across one or more accounts of a user disclosed. One example method may include retrieving a data set for each of a plurality of accounts from a database, constructing a graph including a plurality of nodes linked together by a multitude of edges, wherein each node identifies a time series value corresponding to one of the accounts, and each edge indicates a time series value of a corresponding set of transactions occurring between a corresponding pair of accounts, determining a plurality of constraints, determining a specified loss function based on the plurality of constraints, back-propagating a derivative of the specified loss function into a deep neural network (DNN) to determine a set of neural network parameters, forecasting, using the DNN, a time sequence for one or more of the nodes and one or more of the edges, and providing the forecasted time sequences to the user.

Hierarchical deep neural network forecasting of cashflows with linear algebraic constraints

Exemplary methods for identifying progenies for use in plant breeding are disclosed. One exemplary computer-implemented method includes accessing a data structure including data representative of a pool of progenies and determining a prediction score for at least a portion of the pool of progenies based on the data included in the data structure. The prediction score indicates a probability of selection of the progeny based on historical data. The method further includes selecting a group of progenies from the pool of progenies based on the prediction score, identifying a set of progenies, from the group of progenies, based on at least one of an expected performance of the group of progenies and at least one factor associated with the set of progenies, the pool of progenies and/or the group of progenies, and directing the set of progenies into a validation phase of a breeding pipeline.

Methods and systems for identifying progenies for use in plant breeding

In one embodiment a finite rank deep kernel learning method includes: receiving a training dataset; forming a plurality of training data subsets from the training dataset; for each respective training data subset of the plurality of training data subsets: calculating a subset-specific loss based on a loss function and the respective training data subset; and optimizing a model based on the subset-specific loss; determining a set of embeddings based on the optimized model; determining, based on the set of embeddings, a plurality of dot kernels; combining the plurality of dot kernels to form a composite kernel for a Gaussian process; receiving live data from an application; and predicting a plurality of values and a plurality of uncertainties associated with the plurality of values simultaneously using the composite kernel.

Extending finite rank deep kernel learning to forecasting over long time horizons

Forecasts are provided based on dynamic model selection for different sets of time series. A model comprises a transformation and a prediction algorithm. Given a time series, a transformation is selected for the time series and a prediction algorithm is selected to make a forecast based on the transformed time series. Sets of time series are distinguished from each other based on diverse sparsities, temporal scales and other time series attributes. A model is dynamically selected based on time series attributes to increase forecasting accuracy and decrease forecasting computation time. The dynamic model selection is based on the creation of a meta-model from historical sets of historical time series.

Method and system of dynamic model selection for time series forecasting

In one embodiment, a device receives path telemetry data for one or more network paths in a network over which traffic for an online application is conveyed. The device computes time series dynamics for the path telemetry data. The device determines a mapping of the time series dynamics to application experience metrics for the online application. The device routes traffic associated with the online application based on the mapping.

Application-aware routing based on path KPI dynamics

Systems and methods for forecasting future values of data streams are disclosed. One example method may include receiving information characterizing each of a plurality of forecasting models, retrieving historical data for each of a plurality of data streams, determining one or more constraints, dynamically selecting one of the plurality of forecasting models for each of the data streams based on accuracy metrics for the forecasting models, estimating cost metrics associated with each forecasting model, dynamically selecting the forecasting model based at least in part on the accuracy metrics, the cost metrics, and the determined constraints, and forecasting a first subsequent value of each data stream using the corresponding selected forecasting model.

Model selection in a forecasting pipeline to optimize tradeoff between forecast accuracy and computational cost

Exemplary methods for use in identifying crosses for use in plant breeding are disclosed. One exemplary method includes selecting a subgroup of potential crosses, based on thresholds associated with population prediction scores for the set of potential crosses. The exemplary method further includes selecting multiple target crosses from the subgroup of potential crosses based on a genetic relatedness of the parents in the subgroup of potential crosses, filtering the target crosses based on a rule (or rules) defining a threshold (or thresholds) for at least one characteristic and/or trait, selecting ones of the filtered target crosses based on risk associated with the selected one of the filtered target crosses, and directing the selected ones of the filtered target crosses into a breeding pipeline, thereby providing crosses to the breeding pipeline based, at least in part, on commercial success of parents included in the selected ones of the filtered crosses.

Growing community of inventors

Sunnyvale, CA, United States of America

Sambarta Dasgupta