Abstract: A decoder deployed in one or more terminals, includes a computer readable storage medium storing program instructions, and a processor executing the program instructions, the processor configured to receiving a noisy message and a noisy hash from the network, searching for a pair of matching candidates for the hash and message from two row spaces of noisy message vectors using a shared secret with an encoder, and outputting, by the decoder, a decoded message if the searching is successful.

Abstract: Methods and systems for targeted advertisement include transmitting a pre-filter to a user device, responsive to contextual information supplied by the user device to determine one or more inferences based on physical browsing information, collected at the user device, in compliance with one or more privacy policies of the user. One or more targeted advertisements are determined, using a processor, based on the one or more inferences. The one or more targeted advertisements are transmitted to the user device.

Abstract: A model requester node, which is an edge node of a cloud computing network, generates a specification of a machine learning model, distributes the specification to a plurality of other edge nodes, and receives replies to the specification from the plurality of other edge nodes. In response to the replies, the model requester node identifies a set of participating edge nodes based on a learning utility and a cost estimate of each of the plurality of other edge nodes. The model requester node then trains the machine learning model, without exchanging training data among the model requester node and the participating edge nodes, by repeatedly: distributing most recent parameters of the machine learning model to the participating edge nodes; receiving updates to the most recent parameters from the participating edge nodes; and establishing new parameters for the machine learning model by aggregating the updates from the participating edge nodes.

Abstract: An encoder includes a computer readable storage medium storing program instructions, and a processor executing the program instructions, the processor configured to generate a key, estimate a network capacity, and encode each bit of the key using a random matrix of a selected rank and the estimated network capacity for secure transmission of the key through a network.

Abstract: An encoder including a computer readable storage medium storing program instructions, and a processor executing the program instructions, the processor configured to generating a message by aggregating a plurality of incoming packets, constructing an encoded message using the message and a random matrix, constructing of a hash using a shared secret, and transmitting the encoded message and the hash to a destination, through a network that performs network coding operations.

Abstract: An encoder including a computer readable storage medium storing program instructions, and a processor executing the program instructions, the processor configured to generate a k-bit key, where k is a positive integer, estimate an upper bound of a number of eavesdropped links, encode each bit of the k-bit key using a random matrix of a selected rank, and transmit the encoded k-bit key through a network that performs linear operations on packets.

Abstract: A training process of a machine learning model is executed at the edge node for a number of iterations to generate a model parameter based at least in part on a local dataset and a global model parameter. A resource parameter set indicative of resources available at the edge node is estimated. The model parameter and the resource parameter set are sent to a synchronization node. Updates to the global model parameter and the number of iterations are received from the synchronization node based at least in part on the model parameter and the resource parameter set of edge nodes. The training process of the machine learning model is repeated at the edge node to determine an update to the model parameter based at least in part on the local dataset and updates to the global model parameter and the number of iterations from the synchronization node.

Abstract: Technical solutions are described for selecting components from multiple vendors for a system. An example computer-implemented method includes receiving, by a processor, an environment map that identifies a first component placeholder of the system. The computer-implemented method further includes identifying, by the processor, that the first component placeholder interfaces with a second component placeholder of the system. The computer-implemented method further includes determining, by the processor, a role assigned to the first component placeholder, where the role indicates a type of data that a first component at the first component placeholder communicates with a second component at the second component placeholder. The computer-implemented method further includes determining, by the processor, a list of components for selecting the first component, where each component from the list of components matches the role assigned to the first component placeholder.

Abstract: A data mining method, system, and non-transitory computer readable medium, include defining a set of filter constraints as a filter function for clustering users' private records of data of a private domain, obtaining a set of data from a public domain by applying the filter function to users' public records of data of the public domain, selecting a subset of the users' public records of data that is common with the users' private records of data, and performing data mining on the selected subset of the users' public records of data in combination with the users' private records of data to match a user of the private domain to public records of the user of the private domain.

Abstract: Embodiments of the present invention may involve a method, system, and computer program product for controlling privacy in a face recognition application. A computer may receive an input including a face recognition query and a digital image of a face. The computer may identify a target user associated with a facial signature in a first database based at least in part on a statistical correlation between a detected facial signature and one or more facial signatures in the first database. The computer may extract a profile of the target user from a second database. The profile of the target user may include one or more privacy preferences. The computer may generate a customized profile of the target user. The customized profile may omit one or more elements of the profile of the target user based on the one or more privacy preferences and/or a current context.

Abstract: Status information associated with a set of cloud servers is received. Each cloud server in the set has an associated cloud agent, and the status information includes available computing resources and associated resource compensation information. Aggregated status information is determined for the one or more cloud servers, and the aggregated status information is forwarded to one or more of the cloud agents. A request for performing a computation is received in which the request includes a request specification. A selected cloud agent associated with one or more cloud servers most likely to accept the request based upon the aggregated status information and request specification is determined.

Abstract: Embodiments of the present invention may involve a method, system, and computer program product for controlling privacy in a face recognition application. A computer may receive an input including a face recognition query and a digital image of a face. The computer may identify a target user associated with a facial signature in a first database based at least in part on a statistical correlation between a detected facial signature and one or more facial signatures in the first database. The computer may extract a profile of the target user from a second database. The profile of the target user may include one or more privacy preferences. The computer may generate a customized profile of the target user. The customized profile may omit one or more elements of the profile of the target user based on the one or more privacy preferences and/or a current context.

Abstract: Examples of techniques for interactive generation of labeled data and training instances are provided. According to one or more embodiments of the present invention, a computer-implemented method for interactive generation of labeled data and training instances includes presenting, by the processing device, control labeling options to a user. The method further includes selecting, by a user, one or more of the presented control labeling options. The method further includes selecting, by a processing device, a representative set of unlabeled data samples based at least in part on the control labeling options selected by the user. The method further includes generating, by a processing device, a set of suggested labels for each of the unlabeled data samples.

Abstract: Examples of techniques for adaptive model training are provided. According to one or more embodiments of the present invention, a computer-implemented method for adaptive model training includes generating, by a processing system, a training instance based at least in part on a plurality of images that match a contextual specification of a target visual domain. The method further includes extracting, by the processing system, objects from one of the plurality of images. The method further includes for each extracted object, generating, by the processing system, a plurality of machine learning model features and label recommendations for a user.

Abstract: Examples of techniques for interactive generation of labeled data and training instances are provided. According to one or more embodiments of the present invention, a computer-implemented method for interactive generation of labeled data and training instances includes presenting, by the processing device, control labeling options to a user. The method further includes selecting, by a user, one or more of the presented control labeling options. The method further includes selecting, by a processing device, a representative set of unlabeled data samples based at least in part on the control labeling options selected by the user. The method further includes generating, by a processing device, a set of suggested labels for each of the unlabeled data samples.

Abstract: Examples of techniques for adaptive object recognition for a target visual domain given a generic machine learning model are provided. According to one or more embodiments of the present invention, a computer-implemented method for adaptive object recognition for a target visual domain given a generic machine learning model includes creating, by a processing device, an adapted model and identifying classes of the target visual domain using the generic machine learning model. The method further includes creating, by the processing device, a domain-constrained machine learning model based at least in part on the generic machine learning model such that the domain-constrained machine learning model is restricted to recognize only the identified classes of the target visual domain. The method further includes computing, by the processing device, a recognition result based at least in part on combining predictions of the domain-constrained machine learning model and the adapted model.

Abstract: Examples of techniques for adaptive model training are provided. According to one or more embodiments of the present invention, a computer-implemented method for adaptive model training includes generating, by a processing system, a training instance based at least in part on a plurality of images that match a contextual specification of a target visual domain. The method further includes extracting, by the processing system, objects from one of the plurality of images. The method further includes for each extracted object, generating, by the processing system, a plurality of machine learning model features and label recommendations for a user.

Abstract: Examples of techniques for adaptive object recognition for a target visual domain given a generic machine learning model are provided. According to one or more embodiments of the present invention, a computer-implemented method for adaptive object recognition for a target visual domain given a generic machine learning model includes creating, by a processing device, an adapted model and identifying classes of the target visual domain using the generic machine learning model. The method further includes creating, by the processing device, a domain-constrained machine learning model based at least in part on the generic machine learning model such that the domain-constrained machine learning model is restricted to recognize only the identified classes of the target visual domain. The method further includes computing, by the processing device, a recognition result based at least in part on combining predictions of the domain-constrained machine learning model and the adapted model.

Abstract: Embodiments of the present invention may involve a method, system, and computer program product for controlling privacy in a face recognition application. A computer may receive an input including a face recognition query and a digital image of a face. The computer may identify a target user associated with a facial signature in a first database based at least in part on a statistical correlation between a detected facial signature and one or more facial signatures in the first database. The computer may extract a profile of the target user from a second database. The profile of the target user may include one or more privacy preferences. The computer may generate a customized profile of the target user. The customized profile may omit one or more elements of the profile of the target user based on the one or more privacy preferences and/or a current context.

Abstract: Monitoring a plurality of machines located in an operating environment. First and second acoustic signal readings and their respective detecting locations are received from a sensing device. First and second acoustic signal spatialization map containing characteristic data signatures for the machines are generated based on the first and second acoustic signal readings. One or more differences are determined that exceed a predetermined threshold value, between corresponding characteristic data signatures in each of the first and second acoustic signal spatialization maps. At least one of the machines that are associated with the determined differences is identified. A corrective action to perform on the machine is identified, based on the determined one or more differences. Commands are transmitted to a corrective action module in the operating environment to cause the corrective action module to perform the corrective action.