A system, method, and computer-readable medium for synthesizing zero-copy sparse matrix factorization operations in heterogeneous compute systems are provided. The system includes a host and an accelerator device. The host device is configured to divide an input matrix into a plurality of blocks which are transferred to a memory of the accelerator device. The host device is also configured to generate at least one index buffer that includes pointers to the block in the accelerator's memory, where each index buffer represents a frontal matrix associated with a matrix decomposition algorithm. The host processor is configured to receive one or more kernels configured to process the index buffer(s) on an accelerator device. The index buffers are processed by the accelerator device and the modified block data is written back to a memory of the host device to generate a factorized output matrix.

Wireless communications and/or systems (e.g., 100) and/or methods (e.g., 200, 300, 400) may be provided for predicting of potential undetected flows in a DPI system using a machine learning (ML) model. The system may include an input packet module which may be configured for verifying packet parameters from a network traffic flow, and a processor which can be configured for processing the extracted parameters to identify whether the network traffic flow is potentially detectable or undetectable using a trained machine learning (ML) model based on at least the extracted parameters and perform DPI processing for the detectable flows. Thus, the system may provide an optimized DPI flow processing for high rate traffic networks with decreasing processing time.

A forecasting apparatus forecasts an event after a predetermined time, based on a current window being a part of time-series data in multidimension. The forecasting apparatus includes a non-linear transformation unit including a matrix for non-linear transformation, an observation matrix, and a seasonality setting unit. The non-linear transformation unit transforms the time-series data of the current window in a part of dimensions that are related to trends and the time-series data of the current window in a part of dimensions that are related to seasonal intensity into latent first data showing the trends and latent second data showing the seasonal intensity. The observation matrix includes a first observation matrix that reproduces the first data to first estimated data of an original number of dimensions, and a second observation matrix that, by use of seasonality information that has been set in the seasonality setting unit, reproduces the second data to second estimated data of an original number of dimensions, and adds the first estimated data and the second estimated data.

Systems and methods are provided for vehicle navigation. In one implementation, a system may comprise at least one processor. The processor may be programmed to receive images representative of an environment of the host vehicle and analyze the images to identify a first object and a second object. The processor may determine a first predefined navigational constraint implicated by the first object and a second predefined navigational constraint implicated by the second object, wherein the first and second predefined navigational constraints cannot both be satisfied, and the second predefined navigational constraint has a priority higher than the first predefined navigational constraint. The processor may determine a navigational action for the host vehicle satisfying the second predefined navigational constraint, but not satisfying the first predefined navigational constraint and, cause an adjustment of a navigational actuator of the host vehicle in response to the determined navigational action.

According to various embodiments, the Query Context Translation Engine identifies a topic of a search query history received during a current user session. The search query history in a first language. The Query Context Translation Engine identifies, in a translation table, target text that corresponds with a query in the search query history, the target text comprising at least one word. The Query Context Translation Engine obtains at least one search result based on a translation of the target text in a second language.

According to the present invention, a method for continuous learning of object anomaly detection and state classification model includes acquiring, by a detection and classification apparatus, information about a medium of anomaly detection from an inspection target; generating, by the detection and classification apparatus, an input value, which is a feature vector matrix including a plurality of feature vectors, from the medium information; deriving, by the detection and classification apparatus, a restored value imitating the input value through a detection network learned to generates the restored value for the input value; determining, by the detection and classification apparatus, whether a restoration error indicating a difference between the input value and the restored value is greater than or equal to a previously calculated reference value; and storing, by the detection and classification apparatus, the input value as normal data upon determining that the restoration error is less than the reference value.

A system and method for the estimation of the cardinality of large sets of transaction trace data is disclosed. The estimation is based on HyperLogLog data sketches that are capable to store cardinality relevant data of large sets with low and fixed memory requirements. The disclosure contains improvements to the known analysis methods for HyperLogLog data sketches that provide improved relative error behavior by eliminating a cardinality range dependent bias of the relative error. A new analysis method for HyperLogLog data structures is shown that uses maximum likelihood analysis methods on a Poisson based approximated probability model. In addition, a variant of the new analysis model is disclosed that uses multiple HyperLogLog data structured to directly provide estimation results for set operations like intersections or relative complement directly from the HyperLogLog input data.

Techniques are provided for generating a set of target hybrid seeds with optimal yield and risk performance, including a server receiving a candidate set of hybrid seeds along with probability of successful yield values, associated historical agricultural data and property information, and selecting a subset of the hybrid seeds that have probability of success values greater than a filtering threshold. The server generates representative yield values for hybrid seeds based on the historical agricultural data and risk values for each hybrid seed. The server generates a dataset of target hybrid seeds for planting based on the risk values, the yield values, and the properties for the target fields. The dataset of target hybrid seeds includes target hybrid seeds that meet a specific threshold for a range of risk values. The server causes display of the dataset of target hybrid seeds including yield values and risk values for the target fields.

A mask estimation apparatus for estimating mask information for specifying a mask used to extract a signal of a specific sound source from an input audio signal includes a converter which converts the input audio signal into embedded vectors of a predetermined dimension using a trained neural network model and a mask calculator which calculates the mask information by fitting the embedded vectors to a mixed Gaussian model.

A technique includes receiving, by a processor, a security alert that is generated in response to one or more events occurring in a computer system. The technique includes applying, by the processor, machine learning to the security alert to predict a probability that the security alert will be escalated to an incident; and displaying an output on a display to guide processing of the security alert based on the predicted probability.