Abstract: A data model computing device receives a first data model with a first set of attributes, a first margin of error, a first set of predictions, and an underlying data set. Subsequently, the data model computing device receives a second data model with a second set of attributes, as the test data for a machine learning module. Based on the first and second data model, the machine learning function generates a second set of predictions and a second margin of error. The data model computing device performs a statistical analysis on the first and second set of predictions and the first and second margin of error to determine if the second set of predictions converge with the first set of predictions and second margin of error is narrower than the first margin of error, to determine if the second data model improves the prediction results of the machine learning module.

Abstract: A method for comparing decision options automatically determines factors for a set of decision options. Each of the factors defines a respective attribute of the set of decision options. The method provides the factors to a graphical user interface. The method receives a user preference value for at least one factor. The method then normalizes an attribute value of each factor. The method calculates a user preference score for each decision option as a function of the user preference value and the normalized attribute value for each factor. The method ranks the set of decision options based on the calculated user preference score for the subject decision option relative to each other decision option of the set of decision options. The method provides a ranked list of the set of decision options to the graphical user interface in response to ranking the set of decision options.

Abstract: Systems and methods are provided for receiving a request for lookalike data, the request for lookalike data comprising seed data and generating sample data from the seed data and from user data for a plurality of users, to use in a lookalike model training. The systems and methods further provide for capturing a snapshot of social graph data for a plurality of users and computing social graph features based on the seed data and the user data for the plurality of users, training a lookalike model based on the sample data and the computed social graph features to generate a trained lookalike model, generating a lookalike score for each user of the plurality of users in the user data using the trained lookalike model, and generating a list comprising a unique identifier for each user of the plurality of users and an associated lookalike score for each unique identifier.

Abstract: Systems and methods generate a risk score for an account event. The systems and methods automatically generate a causal model corresponding to a user, wherein the model estimates components of the causal model using event parameters of a previous event undertaken by the user in an account of the user. The systems and methods predict expected behavior of the user during a next event in the account using the causal model. Predicting the expected behavior of the user includes generating expected event parameters of the next event. The systems and methods use a predictive fraud model to generate fraud event parameters. Generation of the fraud event parameters assumes a fraudster is conducting the next event, wherein the fraudster is any person other than the user. The systems and methods generate a risk score of the next event to indicate the relative likelihood the future event is performed by the user.

Abstract: In one embodiment, a method for machine learning acceleration includes receiving, by a shared controller of a tensor processor cluster that includes multiple tensor processors, a multi-cycle instruction, determining, based on the instruction, a sequence of vector operations to be executed by the tensor processors and address information usable to determine a respective spatial partition of an input tensor on which each tensor processor is to operate when performing each vector operation. The method also includes, for each vector operation in the sequence, generating, based on the address information, a common address offset, relative to a respective base address associated with each tensor processor, at which each tensor processor is to retrieve the respective spatial partition on which the tensor processor is to operate, multicasting the common address offset to the tensor processors, and controlling the tensor processors to execute the vector operation in parallel and in lock step.

Abstract: A method for recommending digital content includes: determining user preferences and a time horizon of a given user; determining a group for the given user based on the determined user preferences; determining a number of users of the determined group and a similarity of the users; applying information including the number of users, the similarity, and the time horizon to a model selection classifier to select one of a personalized model of the user and a group model of the determined group; and running the selected model to determine digital content to recommend.

Abstract: A computing server may receive master data, transaction data, and a process model of a domain. The computing server may aggregate, based on domain knowledge ontology of the domain, the master data and the transaction data to generate a fact table. For example, entries in the fact table may be identified as relevant to the target process model and include attributes and facts that are extracted from master data or transaction data. The computing server may convert the entries in the fact table into vectors. The computing server may identify, based on the vectors, an attribute in the process model as being statistically significant on impacting the process model. For example, a regression model may be used to determine the statistical significance of an attribute on the model process. The computing server may generate an action associated with the attribute to improve the process model.

Abstract: In variants, a method for analog product determination can include: determining functional property feature values for a target and determining variable values for a prototype based on the functional property feature values for the target.

Abstract: Systems and methods for synthetic data generation. A system includes at least one processor and a storage medium storing instructions that, when executed by the one or more processors, cause the at least one processor to perform operations including receiving a continuous data stream from an outside source, processing the continuous data stream in real-time, and using machine learning techniques to generating synthetic data to populate the dataset. The operations also include creating a plurality of bins, wherein the plurality of bins occupy a data range between the determined minimum and maximum values without overlapping; and determining a number of samples within each of the created bin, based on a bin edges, wherein the bin edges are bounds within the data range.

Abstract: Techniques for generating a set of Deep Learning (DL) models are described. An example method includes training an initial set of DL models using the training data, wherein a topology of each of the DL models is determined based on the parameters vector. The method also includes generating a set of estimate performance functions for each of the DL models in the initial set based on the set of edge-related metrics, and generating a plurality of objective functions based on the set of estimated performance functions. The method also includes generating a final DL model set based on the objective functions, receiving a user selection of a selected DL model from the final DL model set, and deploying the selected DL model to an edge device.

Abstract: A method includes receiving information associated with a requested operator. The method further includes, in response to receiving the information, generating, by a processing device executing a machine learning model, an artificial intelligence (AI)-based solution to the requested operator, wherein the AI-based solution comprises a plurality of machine-learning models. The method further includes displaying an option to access the AI-based solution in a marketplace platform. The method further includes receiving information associated with a requested operator, and generating, by a processing device executing a first machine learning model, a skeleton architecture of an AI-based solution to the operator based on the information.

Abstract: Sensor data having values received from several sensors of a mobile device and response data associated with the sensor data may be used in the determination or training of a predictive model. Received sensor data may be input into the predictive model, and the output of the predictive model may be used in the selection and serving of content items to the mobile device. Data to effect presentation of the selected content item may be outputted to the mobile device to effect presentation. In some instances, the predictive model may be updated using the received plurality of values. The updated predictive model may be used in the selection of a subsequent content item for the mobile device. In other implementations, historical sensor data may be used with the set of received sensor data as input for the predictive model.

Abstract: Delusional bias can occur in function approximation Q-learning. Techniques for training and/or using a value network to mitigate delusional bias is disclosed herein, where the value network can be used to generate action(s) for an agent (e.g., a robot agent, a software agent, etc.). In various implementations, delusional bias can be mitigated by using a soft-consistency penalty. Additionally or alternatively, delusional bias can be mitigated by using a search framework over multiple Q-functions.

Abstract: A system that incorporates teachings of the present disclosure may include, for example, a computing device including a memory, a display device and a controller. The controller is programmed to provide to a first server a request for User Interface (UI) device configurations, provide to the first server configuration criteria associated with the request, receive configuration data associated with target UI device configurations from the first server, present a Graphical User Interface (GUI) on the display device based on the configuration data where the GUI displays selections for each of the target UI device configurations, provide a selection from among the selections of the target UI device configurations, and receive provisioning information from the first server that allows for implementing the selection of the target UI device configuration. Other embodiments are disclosed.

Abstract: In a hybrid computing system including at least one analog processor and at least one digital processor an embedded problem is repeatedly run or executed on the analog processor(s) to generate a first plurality of candidate solutions to the computational problem, the candidate solutions are returned to the digital processor(s) which determine a value for at least one statistical feature of the candidate solutions, at least one programmable parameter of the plurality of analog devices in the analog processor(s) is adjusted to at least partially compensate for deviations from an expected value of the at least one statistical feature, the expected value of the at least one statistical feature inferred from the structure of the embedded problem, the embedded problem is again repeatedly run or executed on the analog processor(s) to generate a second plurality of candidate solutions to the computational problem.

Abstract: Disclosed are a method and an apparatus for amplitude estimation of a quantum circuit. The method includes: calculating a first difference value between a current angle upper limit value and a current angle lower limit value corresponding to a to-be-estimated amplitude of a target quantum state, and determining the first difference value as a target difference; determining, a next angle amplification factor and a next flag parameter corresponding to a next iteration step; amplifying the target quantum circuit by the next angle amplification factor; calculating a second difference value between a next angle upper limit value and a next angle lower limit value of the to-be-estimated amplitude, and determining the second difference value as a target difference; and determining, based on an angle upper limit value and an angle lower limit value that reach the precision threshold, a probability estimated value corresponding to a to-be-estimated quantum bit.

Abstract: A method and apparatus for certification of facts introduces a certifier and a fact certificate into the fact-exchange cycle that enables parties to exchange trustworthy facts. Certification is provided to a fact presenter during the first part of the fact-exchange cycle, and verification is provided to the fact receiver during the last part of the cycle. To request a certification, a fact presenter presents the Certifier with a fact. In return, the certifier issues a fact certificate, after which the fact presenter presents the fact certificate to the fact receiver instead of presenting the fact itself. The receiver inspects the received certificate in order to evaluate the fact's validity and trustworthiness. For some facts and notions of verification, the certificate is sufficient and its inspection does not require any communication. For others, the receiver requests a verification service from the Certifier in order to complete the verification.

Abstract: Examples of the disclosure are directed toward generating a causation score with respect to an agent and an outcome, and projecting a future causation score distribution. For example, a causation score may be determined with respect to a hypothesis that a given agent causes a given outcome, and the score may indicate the acceptance of that hypothesis in the scientific community, as described by scientific literature. A future causation score distribution, then, may indicate a probability distribution over possible future causation scores, thereby predicting the scientific acceptance of the hypothesis at some specific date in the future. A future causation score distribution can be projected by first generating one or more future publication datasets, and then determining causation scores for each of the one or more future publication datasets.

Abstract: A system is described that performs training operations for a neural network, the system including an analog circuit element functional block with an array of analog circuit elements, and a controller. The controller monitors error values computed using an output from each of one or more initial iterations of a neural network training operation, the one or more initial iterations being performed using neural network data acquired from the memory. When one or more error values are less than a threshold, the controller uses the neural network data from the memory to configure the analog circuit element functional block to perform remaining iterations of the neural network training operation. The controller then causes the analog circuit element functional block to perform the remaining iterations.

Abstract: An apparatus and a method for performing machine learning by executing steps of: generating a decision tree-based machine learning model based on training data; selecting two variables from a decision tree of the machine learning model and determining a correlation between two selected variables; and performing the machine learning based on determined correlation are provided.