Abstract: One or more systems, devices, computer program products and/or computer-implemented methods of use provided herein relate to incorporating explainability constraints into machine learning model training. A computer-implemented system can comprise a memory that can store computer executable components. The computer-implemented system can further comprise a processor that can execute the computer executable components stored in the memory, wherein the computer executable components can comprise a training component that can train a machine learning model using an objective function that can be modified to incorporate an explainability metric for the machine learning model in addition to a model performance metric of the machine learning model.

Abstract: Quantum code-related entities are obtained and one or more of a set of one or more trained neural network models are selected for inferencing based on the quantum code-related entities. The inferencing is performed using the submitted quantum code-related entities and the selected one or more trained neural network models, and a result of the inferencing operation is returned.

Abstract: A method of optimizing an iterative process defines a set of trainable parameters and a differentiable gating function to be applied to each parameter in the set of trainable parameters. A trainable model of the iterative process is built, wherein the iterative process is modified by using the value of the differentiable gating function applied to the parameters to compute a weighted sum of internal variables of the iterative process before and after each iteration. A machine learning-based optimization of the trainable model of the iterative process determines a subset of iterations of the iterative process to perform. The subset of iterations is determined such that an accuracy and a number of active iterations of the iterative process are jointly optimized. The method processes only the subset of the iterations to perform the iterative process. The method is applied to optimize the layered belief propagation algorithm for LDPC decoding.

Abstract: A decision support system and method for an industrial plant is configured and operates to: obtain a causal graph modeling causal assumptions relating to conditional dependence between variables in the industrial plant; obtain observational data relating to operation of the industrial plant; and perform causal inference using the causal graph and the observational data to estimate at least one causal effect relevant for making decisions when operating the industrial plant.

Abstract: A system for dynamically analyzing and configuring nodes of a distributed network leveraging photonic quantum computing is provided. In particular, the system may be configured for identifying one or more nodes of a distributed register network, extracting metadata from the one or more nodes, monitoring the one or more nodes to detect changes to the one or more nodes, determining, via a deep learning network, anomalies associated with the one or more nodes based on the changes, determining target metrics and factors associated with the anomalies, creating simulation test scenarios based on the target metrics, factors, and anomalies, wherein the simulation test scenarios are associated with a plurality of configurations of the one or more nodes, executing the simulation test scenarios in parallel, via a quantum computer, and determining an optimal configuration for the one or more nodes from the plurality of configurations based on executing the simulation test scenarios.

Abstract: In some implementations, a system may receive driving data indicating, for each of a plurality of drivers, driving behavior measured during vehicle operation and importance values corresponding to driving categories associated with the driving behavior. The system may determine driver clusters. The system may receive vehicle data indicating, for each of a plurality of vehicles, performance values corresponding to the driving categories. The system may generate a data structure indicating a degree of association between a vehicle and each driver cluster based on the driving behavior, importance values, and performance values. The system may receive, for a user, driver preference data including data measured during vehicle operation and an importance value for each driving category. The system may determine a driver cluster to which the user belongs based on the driver preference data. The system may determine compatible vehicles based on the driver cluster and the data structure.

Abstract: An analogue arithmetic unit according to an embodiment includes a plurality of input terminals, a voltage/current conversion circuit, a current addition circuit, a current/voltage conversion circuit, and a division circuit. The voltage/current conversion circuit exponentially converts each input voltage applied to each of the plurality of input terminals and outputs the converted input voltage as a current. The current addition circuit obtains a sum of the currents converted by the voltage/current conversion circuit. The current/voltage conversion circuit converts each of the currents and the sum of the currents into voltages. The division circuit calculates a ratio of the voltage obtained by converting each of the currents to a total voltage obtained by converting the sum of the currents.

Abstract: Systems/techniques that facilitate explainable deep interpolation are provided. In various embodiments, a system can access a data candidate, wherein a set of numerical elements of the data candidate are missing. In various aspects, the system can generate, via execution of a deep learning neural network on the data candidate, a set of weight maps for the set of missing numerical elements. In various instances, the system can compute the set of missing numerical elements by respectively combining, according to the set of weight maps, available interpolation neighbors of the set of missing numerical elements.

Abstract: A method includes receiving input data that is used by a machine learning model to generate a first inference. The machine learning model was previously trained using a set of training data. The method also includes generating a set of modified input data based on the input data. The set of modified input data is generated without using the training data. The method further includes obtaining, from an explanation system, an explanation array. The explanation system is to generate the explanation array based on the machine learning model and the set of modified input data.

Abstract: A method includes receiving a time series anomaly detection query from a user and training one or more models using a set of time series data values. For each respective time series data value in the set, the method includes determining, using the trained models, an expected data value for the respective time series data value and determining a difference between the expected data value and the respective time series data value. The method also includes determining that the difference between the expected data value and the respective time series data value satisfies a threshold. In response to determining that the difference between the expected data value and the respective time series data value satisfies the threshold, the method includes determining that the respective time series data value is anomalous and reporting the anomalous respective time series data value to the user.

Abstract: Systems and methods relating to an online system for allowing interactions between online users. The system continuously tracks each user's interactions with different types and categories of content as well as with other users. Based partially on each user's tracked interactions, a probability based model of preferences, interests, and needs is created for each user. The probability based model is then used to predict a user's compatibility with other users and to recommend content. Based on predicted compatibilities, communities and groups of users are recommended to each user. To account for changing user profiles, each user's profile is periodically assessed against the profiles of other users for compatibility. Similarities in interests, preferences, and needs between users, as well as peer rankings, are taken into account when assessing compatibility between users and user groups and communities.

Abstract: Embodiments of the present invention provide computer-implemented methods, computer program products and computer systems. Embodiments of the present invention can, identify a plurality of data variables within a multivariate event dataset. Embodiments of the present invention can then formalize a causal inference between at least two identified data variables within the multivariate event dataset and generate a structural framework of an average effect value for the multivariate event dataset based on the formalization of the causal inference of the identified data variables. Embodiments of the present invention can then calculate an inverse propensity score for the generated structural framework of the average effect based on a type of identified variable, a predetermined time associated with the identified variable, and a causal connection strength between the identified variables.

Abstract: A quantum computation unit, a classical computation unit, and a management unit are provided to learn distribution of a quantum dataset. A structure of a generator is determined from multiple samples of latent variables and values of quantum circuit parameters. A ground cost and a gradient of the ground cost are calculated from the generator and the quantum dataset. An optimal transport loss and a gradient of the optimal transport loss are calculated using the ground cost and the gradient of the ground cost. An updating process is executed that updates the quantum circuit parameter using the gradient of the optimum transport loss, thereby reducing the optimum transport loss. The updating process is repeated until the optimum transport loss converges.

Abstract: A method performed by a processing system including at least one processor includes identifying an insufficiency in a representation of a subpopulation in training data for a machine learning model, generating simulated data to mitigate the insufficiency in the representation, and training the machine learning model using an enhanced training data set that includes the training data and the simulated data to produce a trained machine learning model. In some examples, the generating and the training may be repeated in response to determining that an output of the trained machine learning model still reflects the insufficiency in the representation of the subpopulation or reflects an insufficiency in a representation of another subpopulation. In other examples, the simulated data may be stored for future reuse.

Abstract: There is provided a system and method of automated design space determination for deep neural networks. The method includes obtaining a teacher model and one or more constraints associated with an application and/or target device or process used in the application configured to utilize a deep neural network; learning an optimal architecture using the teacher model, constraints, a training data set, and a validation data set; and deploying the optimal architecture on the target device or process for use in the application.

Abstract: The present disclosure relates to an improved machine learning-based recommender system and method for cold-start predictions on an ecommerce platform. The improved system predicts user-item interactions with respect to cold-start items in which only side information is available. Item representations generated by an item neural network encoder from item side information are shared with a user neural network. The item representations are used, along with user feedback history, to generate user representations. Specifically, a weight matrix in the first layer of the user neural network encoder is fixed with the shared item embeddings. The effect of this is that, when the user neural network encoder is applied to an input user-item interaction vector, the output of the first layer of the user neural network encoder is a function of the item representations of the items for which the user provided positive feedback.

Abstract: Multilingual neural machine translation systems having monolingual adapter layers and bilingual adapter layers for zero-shot translation include an encoder configured for encoding an input sentence in a source language into an encoder representation and a decoder configured for processing output of the encoder adapter layer to generate a decoder representation. The encoder includes an encoder adapter selector for selecting, from a plurality of encoder adapter layers, an encoder adapter layer for the source language to process the encoder representation. The decoder includes a decoder adapter selector for selecting, from a plurality of decoder adapter layers, a decoder adapter layer for a target language for generating a translated sentence of the input sentence in the target language from the decoder representation.

Abstract: Disclosed is a Bayesian classification recognition system based on an industrial PaaS platform, comprising: an IaaS infrastructure service layer, a G-PaaS graph neural network processing layer, an O-PaaS docking service layer and an SaaS system application layer. The G-PaaS graph neural network processing layer is configured for point cloud feature generation, point cloud feature learning, point cloud structure estimation and point cloud model classification; and the recognition accuracy of a workpiece point cloud model is improved through the Bayesian classification recognition system based on the industrial PaaS platform.

Abstract: Systems, devices, and methods are provided for training and/or inferencing using machine-learning models. In at least one embodiment, an identity embedding and a content embedding are extracted from training media data, such as training audio data and/or training visual data. The identity embedding may encode information relating to a person's voice and/or visual properties. A decoder may be trained to create synthetic media, such as audio and visual content, based on identity and content embeddings.

Abstract: Certain aspects of the present disclosure provide techniques for parameterized activation functions. Input data is processed with at least one layer of the neural network model comprising a parameterized activation function, and at least one trainable parameter of the parameterized activation function is updated based at least in part on output from the at least one layer of the neural network model. The at least one trainable parameter may adjust at least one of a range over which the parameterized activation function is nonlinear or a shape of the parameterized activation function, and/or may adjust a location of at least one pivot of the parameterized activation function.