Abstract: Disclosed are techniques for generating features to train a predictive model to predict a customer lifetime value or churn rate. In one embodiment, a method is disclosed comprising receiving a record that includes a plurality of fields and selecting a value associated with a selected field in the plurality of fields. The method then queries a lookup table comprising a mapping of values to aggregated statistics using the value and receives an aggregated statistic based on the querying. Next, the method generates a feature vector by annotating the record with the aggregated statistic. The method uses this feature vector as an input to a predictive model.

Abstract: Techniques are disclosed herein for improving the accuracy of test data obtained outside of a clinical setting. Using the technologies described herein, different techniques can be utilized to analyze, score and adjust test data associated with one or more “at home” tests. In some examples, computing systems are utilized to generate quality scores indicating the accuracy of the test data associated with a particular biomarker. In other examples, an authorized user, such as a data manager can analyze the test data utilizing a user interface to generate scores and/or adjust the test data.

Abstract: In some implementations, a regression prediction platform may obtain one or more feature sets related to an impending code change, wherein the one or more feature sets may include one or more features related to historical code quality for a developer associated with the impending code change or a quality of a development session associated with the impending code change. The regression prediction platform may provide the one or more feature sets to a machine learning model trained to predict a risk associated with deploying the impending code change based on a probability that deploying the impending code change will cause breakage after deployment and/or a probability that the impending code change will cause support issues after deployment. The regression prediction platform may generate one or more recommended actions related to the impending code change based on the risk associated with deploying the impending code change.

Abstract: Debaters encode the logical structure of their arguments in a framework designed so that these arguments can be processed by a computer. A system uses this structure to algorithmically identify logical contradictions and disputed points—places where one debater's views would make another debater's views obviously impossible. The disputed points are represented as short questions whose answers distinguish which debater is right. An end user is presented with these questions, and the user's answers are used to filter out debaters whose views are impossible. The user doesn't need to actually read the debaters' arguments, but can find out which views are right just by answering the questions that have been presented.

Abstract: A method may include receiving an indication that a current instance of an agricultural operation has begun in a geographic area; receiving task characteristics of the agricultural operation; generating a first task data structure including the task characteristics, a field identifier associated with the geographic area, and a task identifier; storing the first task data structure in a database as associated with a first time period; accessing a second task data structure for the agricultural operation associated with a second time period, the second time period being before the first time period; inputting the first and second task data structures into a difference model; receiving an output from the difference model identifying a difference for a first characteristic of the task characteristics between first task data structure and second task data structure; generating a user interface with the identified difference; and presenting the user interface on a computing device.

Abstract: Example implementations described herein involve systems and methods for efficient learning for mixture of domains which can include applying a clustering technique to a set of data comprised of multiple domains to obtain an initial domain separation of the set of data into one or more clusters; training one or more experts associated with each of the one or more clusters based on the initial domain separation where each expert corresponds with one domain of the multiple domains; inputting all data points to the one or more experts for refining each of the one or more clusters using expert output probabilities; retraining the one or more experts based on the refined one or more clusters; and training a gating mechanism to route an input to an appropriate expert of the one or more experts based on the refined one or more clusters.

Abstract: The present technology relates to an information processing system, an information processing method, and a program that are capable of facilitating the understanding of an explanation regarding AI processing. The information processing system includes an evaluation portion that evaluates multiple choices according to two or more evaluation criteria based on parameters obtained during a process or a result of machine learning, and an explanation portion that generates explanatory text regarding each of the choices by using a phrase associated with a corresponding one of the evaluation criteria. For example, the present technology is applicable to hardware and software performing various processes with the use of AI.

Abstract: Methods and systems are described herein for generating a trained Bayesian Hierarchical model from low signal datasets. The disclosed approach utilizes data from alternative segments as a baseline to train the Bayesian Hierarchical model. In some embodiments, the disclosed approach may supplement segment-specific features from another dataset. In some embodiments, inputs for prior distributions may be received from an expert and modified based on the model specification. In one example, the disclosed approach may be used to model probability of default for companies in a low-default segment like Energy portfolio. In this example, data from other commercial and industrial segments is used to form a baseline in the Bayesian Hierarchical model. Further, dataset containing segment-specific features for Energy is supplemented to the training dataset.

Abstract: A method, system and computer program product for assisting users in creating quantum circuits. Quantum logic gates of a quantum circuit (being designed by a user) that are to be grouped together are identified. Upon identifying such quantum logic gates, those quantum logic gates are grouped together into a grouped set of quantum logic gates. Upon forming a grouped set of quantum logic gates, filters are displayed in connection with the quantum circuit being designed by the user, where each of these filters is configured to display particular information (e.g., which gates form the grouped set of quantum logic gates, illustrate how the grouped set of quantum logic gates is transpiled to a hardware system) regarding the grouped set of quantum logic gates. For example, such filters may be visually displayed on the top of the quantum circuit being built by the user in a graphical user interface.

Abstract: According to an embodiment of the present invention, a method, system, and computer program product for reducing and performing quantum circuits. The Embodiment may include receiving, by a classical computer, a quantum circuit comprising a CZ layer and a CNOT layer. The Embodiment may include creating, by a classical computer, a modified quantum circuit based on the CZ layer and the CNOT layer, wherein the modified quantum circuit includes phase gates with CNOT gates that perform similar functions of the CZ gates in the CZ layer. The embodiment may include performing, on a quantum computer, the modified quantum circuit. The embodiment may reduce the depth of a quantum circuit, thereby enabling faster and more accurate computation of the quantum circuit.

Abstract: A method for solving an integer or a mixed-integer programming problem may include: (a) obtaining an indication of a quantum Hamiltonian representative of an integer or a mixed-integer programming problem; (b) implementing the quantum Hamiltonian on a quantum optical device, wherein the quantum optical device comprises quantum gates; (c) using the quantum optical device to prepare a quantum state of the system of qumodes; (d) performing a measurement of the system of the qumodes; and (e) providing a solution of the integer or the mixed-integer programming problem based on the measurement.

Abstract: The present invention relates to an apparatus (12) for processing information by evaluating randomly moving tokens corresponding to particles or quasiparticles, comprising: an input interface (20) for receiving an input signal; a carrier (22) for supporting a plurality of pathways (30) for the randomly moving tokens and at least one join (32) for connecting two pathways and for permitting a passing of a first token in a first pathway of the two connected pathways through the join upon presence of a second token from a second pathway of the two connected pathways in the join; an insertion unit (24) for inserting tokens into the pathways based on the input signal; an excitation unit (26) for applying a stimulus to a token in at least one pathway of the plurality of pathways, said stimulus acting to increase the random movement of the token in a direction substantially parallel to the carrier; and an output unit (28) for providing an output signal based on a location of at least one token after a predefined or d

Abstract: The disclosure is directed towards mitigation of qubit crosstalk-induced errors within a quantum computing system (QCS). A compensating signal is provided to one or more qubits. The compensating signal at least partially “cancels-out” (e.g., compensates for) the crosstalk between pairs of qubits. Such crosstalk-induced errors may include leakage of a qubit's quantum state out of the quantum system's computational subspace. Thus, the embodiments may be employed to decrease quantum computational errors occurring from a qubit transitioning (or leaking) to an excited state that is not within the quantum system's computational subspace.

Abstract: Systems, computer-implemented methods or computer program products to facilitate mitigating quantum errors associated with one or more quantum gates. A noise modeling component can generate a sparse error model of noise associated with one or more quantum gates; employ the sparse error model; and draw samples from an inverse noise model. An insertion component can insert the samples to mitigate errors associated with the one or more quantum gates. The insertion component can reduce the noise by running circuit instances augmented with samples from the inverse noise model. The noise modeling component includes a noise shaping component that can shape the noise affecting one or more quantum gates by twirling to form a Pauli channel.

Abstract: An N-value prediction apparatus according to an embodiment of the present invention includes a hypothetical learning data augmentation unit, based on an actual N-value measured at an actual location according to the Standard Penetration Test through drilling investigation, generating at least one of hypothetical N-values corresponding to a preset hypothetical point based on the actual location, an N-value prediction model learning unit learning ground characteristic data corresponding to each of the actual location and the hypothetical point by artificial intelligence, the ground characteristic data including the actual N-value and the hypothetical N-values, and an N-value prediction result calculation unit predicting an N-value at an arbitrary prediction target point by using an N-value prediction model generated by artificial intelligence learning executed in the N-value prediction model learning unit.

Abstract: The disclosure relates to a method for generating a custom predictive model, the method comprising receiving a plurality of datasets; identifying a plurality of features affecting prediction of a predictive model; determining an importance score for each of the plurality of features; determining a probability of prediction for each dataset from the plurality of the datasets based on one or more features from the plurality of features for the prediction of each dataset, and respective importance scores for the one or more features; and training a custom predictive model using the plurality of datasets with the probabilities, the respective features, and the respective importance scores.

Abstract: Some embodiments select a machine learning model training duration based at least in part on a fractal dimension calculated for a training data dataset. Model training durations are based on one or more characteristics of the data, such as a fractal dimension, a data distribution, or a spike count. Default long training durations are sometimes replaced by shorter durations without any loss of model accuracy. For instance, the time-to-detect for a model-based intrusion detection system is shortened by days in some circumstances. Model training is performed per a profile which specifies particular resources or particular entities, or both. Realistic test data is generated on demand. Test data generation allows the trained model to be exercised for demonstrations, or for scheduled confirmations of effective monitoring by a model-based security tool, without thereby altering the model's training.

Abstract: Systems and methods for data augmentation are described. Embodiments of the present disclosure receive a dataset that includes a plurality of nodes and a plurality of edges, wherein each of the plurality of edges connects two of the plurality of nodes; compute a first nonnegative matrix representing a homophilous cluster affinity; compute a second nonnegative matrix representing a heterophilous cluster affinity; compute a probability of an additional edge based on the dataset using a machine learning model that represents a homophilous cluster and a heterophilous cluster based on the first nonnegative matrix and the second nonnegative matrix; and generate an augmented dataset including the plurality of nodes, the plurality of edges, and the additional edge.

Abstract: An electronic device and a method for determining scenario data of a self-driving car are provided. The method includes: obtaining training scenario data by using scenario data, a loss function and a self-driving program module; training an encoding module and a decoding module by using the training scenario data, and generating a scenario space by using the trained encoding module; obtaining a monitoring module by using the scenario space; and executing the monitoring module to determine whether current scenario data belongs to an operational design domain by using the current scenario data and the trained encoding module.

Abstract: Provided are a method and an apparatus for operating a learning model. A method for operating a learning model according to one embodiment of the present disclosure comprises selecting at least one training data between previous training data and new training data and learning a previous learning model anew using the at least one selected training data.

Abstract: An apparatus and method for operation of a supervisory platform, the apparatus including at least a processor; and a memory connected to the at least a processor, the memory containing instructions configuring the at least a processor to receive a user profile from a user utilizing a smart assessment, generate a tending program as a function of the user profile utilizing a tending program machine learning model, wherein generating the tending program includes training the tending program machine learning model using training data, the training data includes user profile data correlated to tending program data and generate a tending outcome as a function of the tending program and the user profile.

Abstract: A non-transitory computer-readable recording medium stores a program for causing a computer to execute a process, the process include identifying frequency components stronger than a predetermined reference among frequency components of time-series data, calculating values that indicate a relationship between one or more parameters used when generating a plurality of time-series features of the time-series data and periods having the identified frequency components, as features for the parameters, executing training of a first machine learning model by using importance of each of the time-series features on prediction that uses the time-series features and the features for each of the parameters to predict the importance of the time-series features from the features for each of the parameters, the time-series features being generated based on the parameters, and predicting importance of time-series features for new time-series data by using the trained first machine learning model.

Abstract: In a compatibility evaluation device, the acquisition means acquires outputs of a first predictor and a second predictor for evaluation data. The index determination means determines a generalized backward compatibility index defined by a combination of a plurality of relational expressions representing relationship between the output of the first predictor and the output of the second predictor. The calculation means calculates a score indicating compatibility of the first predictor and the second predictor, using the output of the first predictor, the output of the second predictor, and the generalized backward compatibility index.

Abstract: A method for detecting and/or preventing overfitting in training of deep learning and neural network models. The method has a classifier-training method, an overfitting-detection method, and an overfitting-prevention method. The classifier-training method trains one or more classifiers using training histories and labels of one or more trained machine-learning (ML) models. The overfitting-detection method uses the trained classifiers based on the training history such as validation losses of a trained target ML model to identify an overfitting status of the trained target ML model. The overfitting-prevention method is performed during the training of a target ML model and uses the trained classifiers based on the training history of the target ML model to identify and preventing overfitting of the target ML model.

Abstract: A solution for generating a negative sample pair of a contrastive learning model is provided. In one method, a first data segment is obtained from a first data sequence in a plurality of data sequences for training the contrastive learning model, and a second data segment is obtained from a second data sequence in the plurality of data sequences. A data frame is selected from a further data sequence than the second data sequence in the plurality of data sequences. A third data segment is generated based on the second data segment and the data frame. A negative sample pair for training the contrastive learning model is determined based on the first data segment and the third data segment. Therefore, richer semantic information can be introduced into negative sample pairs in terms of appearance, and further the accuracy of the contrastive learning model can be improved.

Abstract: A data processing method applied to a database system is disclosed. The method includes: obtaining a model training request, where the model training request includes a plurality of training samples and a model training policy, and the plurality of training samples is grouped into N training sample groups; generating an execution plan of the model training policy and an estimated execution cost of the execution plan executed by the database system; obtaining, based on the estimated execution cost, M training sample groups in the N training sample groups; training a to-be-trained model in parallel by using the M training sample groups, to obtain M pieces of parameter update data; and updating the to-be-trained model based on the M pieces of parameter update data, to obtain a trained model. This method reduces time overheads of the model training.

Abstract: A method for managing training data of a biosignal analysis model includes the steps of: converting data on at least one of a plurality of leads into augmented data associated with a specific lead among the plurality of leads; and training an analysis model for determining arrhythmia on the basis of data on the specific lead, using the augmented data and the data on the specific lead as training data.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for providing a machine learning model that is trained to perform a machine learning task. In one aspect, a method comprises receiving a request to train a machine learning model on a set of training examples; determining a set of one or more meta-data values characterizing the set of training examples; using a mapping function to map the set of meta-data values characterizing the set of training examples to data identifying a particular machine learning model architecture; selecting, using the particular machine learning model architecture, a final machine learning model architecture for performing the machine learning task; and training a machine learning model having the final machine learning model architecture on the set of training examples.

Abstract: A method for monitoring performance of a ML system includes receiving a data stream via a processor and generating a first plurality of metrics based on the data stream. The processor also generates input data based on the data stream, and sends the input data to a machine learning (ML) model for generation of intermediate output and model output based on the input data. The processor also generates a second plurality of metrics based on the intermediate output, and a third plurality of metrics based on the model output. An alert is generated based on at least one of the first plurality of metrics, the second plurality of metrics, or the third plurality of metrics, and a signal representing the alert is sent for display to a user via an interface.

Abstract: A substantial learning curve is required to construct integration processes in an integration platform. This can make it difficult for novice users to construct effective integration processes, and for expert users to construct integration processes quickly and efficiently. Accordingly, embodiments for building and operating a model to predict next steps, during construction of an integration process via a graphical user interface, are disclosed. The model may comprise a Markov chain, prediction tree, or an artificial neural network (e.g., graph neural network, recurrent neural network, etc.) or other machine-learning model that predicts a next step based on a current sequence of steps. In addition, the graphical user interface may display the suggested next steps according to a priority (e.g., defined by confidence values associated with each step).

Abstract: Disclosed are various example embodiments which may be configured to: receive, from a distributed node, local dataset information comprising characteristics of a local dataset of the distributed node, assign a score to the distributed node and/or determine whether the distributed node is a potential malicious distributed node based on the local dataset information, determine whether to select the distributed node for training a local model for managing a network in a federated learning mechanism based on the score assigned to the distributed node and/or whether the distributed node is a potential malicious distributed node, and send, to the distributed node, an indication as to whether the distributed node has been selected for training a model for managing a network in a federated learning mechanism.

Abstract: Machine learning (ML) processes ranks data instances to determine when to adjust parameters of a machine learning model. Data instances are ranked by receiving data instances. Further, ranked instances are determined based on the data instances and a machine learning model. A metric is determined based on the ranked instances. An adjusted machine learning model is generated by adjusting one or more parameters of the machine learning model based on the metric.

Abstract: A method comprising: determining at a first device a requirement for a spatial distribution associated with training samples for a machine learning model, wherein the machine learning model is used in solving a positioning task; performing at least one of: transmitting, to the second device, information indicating the requirement for training the machine learning model, or training the machine learning model by using a training dataset comprising training samples satisfying the requirement.

Abstract: A solution for generating a positive sample pair of a contrastive learning model is provided. In one method, a first data segment and a second data segment are respectively obtained from a first data sequence in plurality of data sequences for training the contrastive learning model, the first data sequence comprising a plurality of data frames. A data frame is selected from a second data sequence in the plurality of data sequences. A third data segment is generated based on the second data segment and the data frame. A positive sample pair for training the contrastive learning model is determined using the first data segment and the third data segment. In this way, positive samples that are more difficult to distinguish can be provided, thereby increasing the accuracy of the contrastive learning model and improving the training efficiency and accuracy of downstream models.

Abstract: A method, apparatus, device, and medium for managing a contrastive learning model are provided. In one method, in a first training phase, a first contrastive learning model is generated by training the contrastive learning model with a first training sample set, a negative sample pair of the first training sample set comprising only data segments from different data sequences. In a second training phase, a second contrastive learning model is generated by training the first contrastive learning model with a second training sample set, a negative sample pair in the second training sample set comprising data segments from the same data sequence. Knowledge in terms of the appearance of the samples can be fully obtained in the first training phase, and knowledge in terms of the appearance and dynamics of the samples can be fully obtained in the second training, with example implementations of the present disclosure.

Abstract: A contextual awareness subsystem extracts information about the environment around a device from audio data. An adaptive learning model may be applied to the contextual information to generate a recommendation of a change to the configuration of the device. The recommendation may be automatically implemented or presented to the user for verification.

Abstract: A method is provided for debiasing machine learning models. The method includes obtaining (i) an initial model that is a trained and tree-based machine learning model and (ii) a minimum acceptable threshold accuracy, for (iii) one or more protected classes. The initial model demonstrates adverse impact on one or more protected classes. The method includes identifying branches of the initial model to prune, based on the branches' impact on one or more protected classes. The method includes applying a pruning algorithm to prune the branches of the initial model to generate one or more forest models, such that (i) predictive accuracy of the one or more forest models is above the minimum threshold accuracy, and (ii) the one or more forest models are less discriminatory than the initial mode.

Abstract: Computer-implemented method for training a machine learning system to denoise a provided input signal. The method includes: providing a first input signal and a first value to a first part of the machine learning system, wherein the first input signal characterizes a noisy signal; determining, by the first part, a first output signal for the first input signal and the first value; determining, by a second part of the machine learning system, a second value based on the first output signal, wherein the second value characterizes a probability of the first output signal to characterize a noisy signal; determining, by the second part, a third value based on a supplied second input signal, wherein the second input signal characterizes a non-noisy signal and wherein the third value characterizes a probability of the second input signal to characterize a non-noisy signal; training the machine learning system.

Abstract: An apparatus includes circuitry configured to: receive a request from a radio access network algorithm to determine whether there is a distribution shift related to a temporal characteristic of a cell of a communication network; request data from a radio access network node or a controller platform related to the temporal characteristic; receive the requested data related to the cell from the radio access network node or the controller platform; determine whether there is a distribution shift related to the temporal characteristic; in response to determining that there is a distribution shift, select a learning type for an update to a model; and update the model such that, when the model is provided to an inference server, causes the radio access network algorithm to use the updated model to perform at least one action to optimize the performance of the radio access network node or other radio access network node.

Abstract: Methods and systems are provided for monitoring a physical asset, which includes receiving or collecting time-series data related to operation or status of the physical asset; identifying a time period when the physical asset is experiencing a change in operational state; extracting time-series data corresponding to the time period as event data; generating label data that classifies or characterizes the event data as pertaining to a particular type of event; saving the event data and the corresponding label data in a data repository; and using the event data and label data stored in the data repository to train or update a machine learning system to detect the occurrence of events that are similar to the event types of the labeled event data stored in the data repository from time-series data generated by the physical asset or by another physical asset that operates in a similar manner to the physical asset.

Abstract: A learning device acquires learning data of a model predicting a label of input data including an adversarial example. The learning device performs learning of the model using a loss function that flattens a loss landscape with respect to a parameter by adding noise in which KL divergence of a loss value in the model becomes maximum to the parameter of the model and learning data including the adversarial example when the noise is added to the parameter of the model and when the noise is not added.

Abstract: A method comprises receiving logistics operation order data, wherein the logistics operation order data identifies at least one logistics operation to be performed. The logistics operation order data is analyzed using one or more machine learning algorithms. Based at least in part on the analyzing, a logistics provider to perform the at least one logistics operation is predicted.

Abstract: A management apparatus includes a controller configured to present a power supply service available to a vehicle to be power supplied based on a status of the power supply service, and accept a reservation for use of the power supply service from a user of the vehicle to be power supplied.

Abstract: The present disclosure relates to a travel scheduling system and method thereof, and more particularly, to a travel scheduling system in which a user can easily and quickly establish a travel schedule by displaying a travel schedule list including a plurality of travel schedules on a terminal of a first user, receiving a selection of any one travel schedule by the first user from among a plurality of travel schedules from the terminal of the first user, and generating a travel schedule of the first user by copying the travel schedule selected by the first user, and a method thereof.

Abstract: After receiving a new task which specifies traveling from a first source bin to a first target bin, a computer-implemented method can select a task which specifies traveling from a second source bin to a second target bin, calculate a connecting distance to the new task measured from the second target bin to the first source bin, and compare the connecting distance to the new task with connecting distances to candidate tasks of the selected task. Responsive to determining that the connecting distance to the new task is smaller than the connecting distance to one of the candidate tasks, the method can update a successor list associated with the selected task. Responsive to completion of the selected task, the method can select a successor task from the successor list to execute. The successor task has the shortest connecting distance among the candidate tasks of the selected task.

Abstract: Auditing of page profile data on sources is disclosed. Normalized reference location data associated with attributes of an entity location are obtained. Page data from the profile of the entity location on a source is collected. The collected page data is normalized. It is determined whether the page data on the profile of the entity location on the source is valid at least in part by comparing the normalized collected page data against the normalized reference listing data. An audit score is assigned to the profile of the entity location on the source based at least in part on the validity determination. Output is provided based at least in part on the audit.

Abstract: A task management service server capable performing task management for an electronic terminal and the operating method thereof are provided to support an administrator to more easily perform task setting and management for an electronic terminal on a predetermined person to be managed.

Abstract: A computer implemented method comprising displaying a scheduling interface including a first work item user interface element corresponding to a first work item and a second work item user interface element corresponding to a second work item; detecting a create dependency interaction; and, in response to detecting the create dependency user interaction, creating a dependency relationship between the first and second work items.

Abstract: A method for presenting data about a construction site on a network of operatively connected computers, comprises the steps of establishing authorized workers at the construction site authorized to complete at least a portion of a pre-task plan which corresponds to the authorized worker, presenting the portion of the pre-task plan to the authorized worker, generating inputs from the authorized worker corresponding to a change in status which are used to create reports, and pushing reports to selected parties on the network based on the inputs from the authorized worker in real-time.

Abstract: Oil field resources are allocated using machine learning and optimization. A well of a set of wells is identified. A user interface is presented to obtain a priority of a set of priorities. A schedule is generated for a set of resources for a job using the priority. A job schedule is presented that includes a set of jobs for the set of wells. The job schedule is generated using the set of priorities. An update to the priority is obtained. An updated job schedule is presented based on the update to the priority.