Abstract: Systems and methods are disclosed for parallelizing service function chains.

Abstract: Systems and methods for computing a causal uplift in performance of an output action for one or more treatment actions in parallel are described herein. In an embodiment, a server computer receives interaction data for a particular period of time which identifies a plurality of users and a plurality of actions that were performed by each user of the plurality of users through a particular graphical user interface during the particular period of time. The server computer uses the interaction data to generate a feature matrix of actions for each user, and a set of confounding variables included to minimize spurious correlations. The feature matrix is then used to train a machine learning system, using data identifying a user's performance or non-performance of each action as inputs and data identifying performance or non-performance of a target output action as the output.

Abstract: A processing system may obtain an intent comprising a desired network state for a telecommunication network, determine an existing network state, where the desired network state and the existing network state are defined in accordance with a graph representing the telecommunication network, where the graph includes objects and relationships between the plurality of objects, and where each of the objects includes one or more properties. The processing system may compose a strategy that includes a plurality of policies to obtain the desired network state from the existing network state, where each of the policies comprises an action that is to be implemented in the telecommunication network, where the composing comprises performing a state exploration process over the graph, and where each state transition of the process corresponds to one of the policies. The processing system may then implement the strategy to place the telecommunication network in the desired network state.

Abstract: Implementations described herein relate to methods, systems, and computer-readable media for automated generation and use of a machine learning (ML) model to provide recommendations. In some implementations, a method includes receiving a recommendation specification that includes a content type and an outcome identifier, and determining model parameters for a ML model based on the recommendation specification. The method further includes generating a historical user feature matrix (FM), generating a historical content feature matrix (FM), and transforming the historical user FM and the historical content FM into a suitable format for the ML model. The method further includes obtaining a target dataset that includes historical results for the outcome identifier for a plurality of pairs of user identifiers and content items of the content type. The method further includes training the ML model using supervised learning to generate a ranked list of content items for each user identifier.

Abstract: Performance issues in a service function chain having a plurality of resources and a plurality of network functions each having a network function queue are diagnosed. Each network function queue is monitored and queueing information for input packets for each of the plurality of network functions is dumped to a data store. Each resource that is under contention is identified as well as which of the network functions is a contender for the resources. A diagnosing algorithm is used to diagnose performance problems and an impact graph for each victim packet is generated. A summary of results as a list of rules is then provided.

Abstract: A system for monitoring marine animals includes a monitoring tag attachable to a marine animal. The monitoring tag includes a processor, a memory coupled to the processor, at least one communication interface, and at least one sensor. The processor is configured to store readings from the at least one sensor in the memory. The system also includes at least one communication receiver having a processor, a memory coupled to the processor, and at least one communication interface. The at least one communication receiver is configured to float in water.

Abstract: In operating an ecommerce marketplace, pre-generated metadata is used to transform images after a request is received, in real-time, and with low latency. To accomplish the high-speed image transformation, an offline annotation of relevant image features is obtained using machine learning and metadata is stored based upon the results. The metadata is then used to perform the high-speed image transformation at request time. The transformations can include image cropping, adjustment in saturation, contrast, brightness, extracting portions of the image, etc. In one example, the metadata includes a bounding box giving coordinates of how to crop an image so that the image can be cropped without analyzing content or context of the image. Instead, to obtain the coordinates of the bounding box, content and context of the image are analyzed in pre-request processing. In this way, when the request is received, the more difficult and time-consuming processing is already completed.

Abstract: To simulate automotive systems, a large number of synthetic data points characterising an aspect of the performance of a target automotive system are generator. In this way, for example, various future scenarios can be simulated and statistically evaluated. A computer-implemented method is provided for training a generative machine learning model, a computer-implemented method for generating synthetic data series using a generative machine learning model, and an associated apparatus. An associated computer program element and computer readable medium are also described.

Abstract: A processing system may obtain an intent comprising a desired network state for a telecommunication network, determine an existing network state, where the desired network state and the existing network state are defined in accordance with a graph representing the telecommunication network, where the graph includes objects and relationships between the plurality of objects, and where each of the objects includes one or more properties. The processing system may compose a strategy that includes a plurality of policies to obtain the desired network state from the existing network state, where each of the policies comprises an action that is to be implemented in the telecommunication network, where the composing comprises performing a state exploration process over the graph, and where each state transition of the process corresponds to one of the policies. The processing system may then implement the strategy to place the telecommunication network in the desired network state.

Abstract: Views may be generated for bias metrics or feature attribution captured in machine learning pipelines. A request to create a view of bias metrics or feature attribution may be received. The bias metrics or feature attribution may have been determined in a machine learning pipeline as part of executing a training job that specified the bias metrics or the feature attribution. A development application may access a data store that stores the bias metrics or the feature attribution determined in the machine learning pipeline. A view based on the bias metrics or feature attribution may be generated and provided.

Abstract: Bias metrics may be captured at different stages for training a machine learning model. A training job may specify bias metrics to capture at multiple different stages of a machine learning pipeline for a feature of a training data set used to train a machine learning model. The training job may be executed and the bias metrics determined at the stages as specified in the training job. The bias metrics for the different stages may be stored.

Abstract: Bias metrics and feature attribution may be monitored for a machine learning model. A request to enable monitoring for bias metrics or feature attribution may be received. Monitoring may be enabled to evaluate respective performance of inferences of a machine learning model according to the enabled bias metrics or feature attribution. If a divergence from reference data is detected, then a notification indicating the divergence may be sent.

Abstract: Feature attribution may be captured as part of a machine learning pipeline. A training job may include a request to determine feature attribution as part of a machine learning pipeline that trains a machine learning model from a training data set. A reference data set for determining the feature attribution of the machine learning model may be identified. The feature attribution may be determined based on the reference data set. The feature attribution of the trained machine learning model may be stored.

Abstract: Performance issues in a service function chain having a plurality of resources and a plurality of network functions each having a network function queue are diagnosed. Each network function queue is monitored and queueing information for input packets for each of the plurality of network functions is dumped to a data store. Each resource that is under contention is identified as well as which of the network functions is a contender for the resources. A diagnosing algorithm is used to diagnose performance problems and an impact graph for each victim packet is generated. A summary of results as a list of rules is then provided.

Abstract: Techniques are described for implementing an out-of-band communication channel used to exchange control channel information in sub-carrier-based optical communication systems. In an example implementation, a transmitter includes a laser operable to supply an optical signal, a digital signal processor operable to supply first electrical signals based on first data input to the digital signal processor and second data input to the digital signal processor, digital-to-analog conversion circuitry operable to output second electrical signals based on the first electrical signals, modulator driver circuitry is operable to output third electrical signals based on the second electrical signals, and an optical modulator operable to supply first and second modulated optical signals based on the third electrical signals. The first modulated optical signal includes a plurality of optical subcarriers carrying user data. The plurality of optical subcarriers also being amplitude modulated to carry control information.

Abstract: Systems and methods for performing impairment compensation in point-to-multi-point communication systems are described. In a data snapshot mode, a hub node can send instructions to each communication node connected to the hub node to send a data snapshot of data being received and processed by the communication nodes at a particular time. In a trench line mode, a hub node sends instructions to each communication node to send trench line data back to the hub node. The hub node uses the data snapshot or trench line data to determine how to tune filter coefficients in the hub node to perform impairment compensation and improve performance of the communication system.

Abstract: The present invention discloses a power control apparatus and method in a wireless communication system. According to the present invention, among FUE units neighboring to an MUE unit of a victim cell, an FUE unit of an aggressor cell may be derived on the basis of information on RSRP of the FUE unit registered in a closed subscriber group, and a threshold interference value of the FUE unit. Also, power of the derived FUE unit of the aggressor cell may be gradually decreased until an interference of the MUE unit of the victim cell becomes greater than a predefined threshold interference value, and an SINR of the FUE unit of the aggressor cell becomes equal to or greater than a target SINR. Accordingly, CCI can be reduced in a heterogeneous network environment, data transmission speed can be improved, and latency time can be reduced by reducing feedback time.

Abstract: A system may assist with checking policy impact in a software-defined infrastructure environment. The system's data analysis may enable it to discover and quantify the impact of policies on software-defined infrastructure objects in the same or different layers.

Abstract: Systems, computer-readable media, and methods are disclosed for parallel data processing for service function chains with network functions spanning multiple servers. An example system includes a first server hosting a first network function of a service function chain, a second server hosting a second network function of the service function chain, a mirror function deployed in a first switch to replicate a plurality of packets received by the system and to send respective copies of the plurality of packets to the first network function and to at least one of the second network function and a third network function of the service function chain, and a merge function deployed in a second switch to merge respective outputs of the first network function and the at least one of the second network function and the third network function.

Abstract: In operating an ecommerce marketplace, pre-generated metadata is used to transform images after a request is received, in real-time, and with low latency. To accomplish the high-speed image transformation, an offline annotation of relevant image features is obtained using machine learning and metadata is stored based upon the results. The metadata is then used to perform the high-speed image transformation at request time. The transformations can include image cropping, adjustment in saturation, contrast, brightness, extracting portions of the image, etc. In one example, the metadata includes a bounding box giving coordinates of how to crop an image so that the image can be cropped without analyzing content or context of the image. Instead, to obtain the coordinates of the bounding box, content and context of the image are analyzed in pre-request processing. In this way, when the request is received, the more difficult and time-consuming processing is already completed.