Abstract: The system gathers a set of biometric data for a first driver of a vehicle. The system classifies the set of biometric data as belonging in one of a plurality of biometric conditions. The plurality of biometric conditions include at least one normal biometric condition representative of a normal, attentive condition and a first abnormal biometric condition representative of an inattentive condition. The system determines a personal threshold between the normal biometric condition and the first abnormal biometric condition for the first driver for a first biometric parameter. In a driving phase, the system monitors biometric data of the first driver to determine whether the first driver has passed from the normal biometric condition to the first abnormal biometric condition according to the personal threshold.

Abstract: The present application relates to apparatus, systems, and methods for grouping data records based on entities referenced by the data records. The disclosed grouping mechanism can include determining a pair-wise similarity between a large number of data records, and clustering a subset of the data records based on their pair-wise similarity.

Abstract: System and methods for machine learning are described. A first input value is obtained. A second input value is also obtained. A decision to use for generating a cycle output is selected based on a randomness factor. The decision is at least one of a random decision or a best decision from a previous cycle. A cycle output for the first and second inputs is generated using the selected decision. The selected decision and the resulting cycle output are stored.

Abstract: A framework for improving data set in an enterprise system for machine learning is provided. In accordance with one aspect, user input of a project update is provided by a user to an enterprise system. A record of the project update is created in the enterprise system based on the user input. The project update provided by the user into the enterprise system is analyzed using a gamification technique. The analysis includes quantifying the user's input of the project update to the enterprise system by assigning points to the user based on the project update provided to the enterprise system. The assigned points are displayed to the user on a user interface of a user device to enable friendly competition with other users which encourages more detailed and frequent project updates to the enterprise system by the user.

Abstract: Systems and arrangements for using temporal analysis to evaluate incidents to determine whether they are likely to cause a significant business impact are provided. Historical data may be analyzed to identify incidents having a significant business impact. The historical data associated with incidents having a significant business impact may be further analyzed to identify a time and/or date at which the incident occurred, as well as the particular system, or the like, impacted by the incident. Normal business hours associated with the system, or the like, may be retrieved and a profile may be generated for the system, or the like. An incident may be received and processed to identify a system, or the like, associated with the incident and profile may be retrieved. The incident data may be compared to the profile to determine whether the incident is likely to cause a significant business impact based, at least in part, on the date and/or time at which it occurred.

Abstract: A system and method for recognizing characters embedded in multimedia content are provided. The method includes extracting at least one image of at least one character from a received multimedia content item; identifying a natural language character corresponding to the at least one image of the at least one character, wherein the identification is performed by a deep content classification (DCC) system; and storing the identified natural language character in a data warehouse.

Abstract: Systems and methods for machine learning and adaptive optimization are provided herein. A method includes continually receiving input that is indicative of client events, including client behaviors and respective outcomes of software trials of a product maintained in a database, continually segmenting open opportunities using the client behaviors and respective outcomes, continually scoring and prioritizing the open opportunities using the client behaviors and respective outcomes for targeting and re-targeting, continually adjusting targeted proposals to open opportunities and sourcing in prospects based on a targeting scheme, continually presenting targeted offers to create expansion opportunities and updating a product roadmap of the product using the open opportunities, the product roadmap including technical specifications for the product.

Abstract: A technique for estimating a set of parameter values for a lumped parameter model of a loudspeaker. The technique includes determining, via a neural network model, a first set of parameter values for the LPM of the loudspeaker based on an audio input signal and a first measured response of a loudspeaker that corresponds to the audio input signal. The technique further includes generating a first LPM response based on the first set of parameter values and comparing the first LPM response to the first measured response of the loudspeaker to determine a first error value. The technique further includes generating, via the neural network model, a second set of parameter values for the LPM of the loudspeaker based on the first error value.

Abstract: A system for prospectively identifying media characteristics for inclusion in media content is disclosed. A neural network database including media characteristic information and feature information may associate relationships among the media characteristic information and feature information. Personal characteristic information associated with target media consumers may be used to select a subset of the neural network database. A first set of nodes, representing selected feature information, may be activated. The node interactions may be calculated to detect the activation of a second set of nodes, the second set of nodes representing media characteristic information. Generally, a node is activated when an activation value of the node exceeds a threshold value. Media characteristic information may be identified for inclusion in media content based on the second set of nodes.

Abstract: Technologies are described herein for identification and presentation of changelogs relevant to a tenant of a multi-tenant cloud service. Change feature extraction is performed on changelogs associated with a tenant of the multi-tenant cloud service to identify features associated with the changelogs. Machine learning based classification can then be performed on the changelogs to classify the changelogs. Misclassification correction might also be performed on the classified changelogs. Machine learning can also be utilized to identify a subset of the changelogs as being relevant to the tenant. A user interface (UI) can then be generated and provided to the tenant that includes the subset of the changelogs. The tenant's interaction with the changelogs presented in the UI can be monitored and data describing the interaction can be used to modify machine learning models utilized for machine learning change classification and for determining the relevance of a changelog to the tenant.

Abstract: In certain embodiments, a method includes formulating an optimization problem to determine a plurality of model parameters of a system to be modeled. The method also includes solving the optimization problem to define an empirical model of the system. The method further includes training the empirical model using training data. The empirical model is constrained via general constraints relating to first-principles information and process knowledge of the system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for predictive modeling for adjusting initial values are disclosed. In one aspect, a method includes the actions of accessing transaction history data that indicates one or more transaction details associated with the transaction, a predicted value, and a final value. The actions further include determining a difference value between the predicted value and the final value. The actions further include generating a predictive model that is trained to estimate. The actions further include receiving one or more transaction details and a predicted value associated with a subsequently received transaction. The actions further include providing the one or more transaction details as input to the predictive model. The actions further include receiving an adjustment value to apply to the predicted value. The actions further include providing, for output, data indicating the adjustment value.

Abstract: Embodiments are directed to identifying active compounds for a targeted medium from a library of compounds. In one scenario, a computer system receives high throughput screening (HTS) data for a subset of compounds that have been HTS-screened. The computer system determines labels for a subset of compounds based on labels identified in the HTS-screened compounds as being part of an active class or part of an inactive class, access chemical features corresponding to the HTS-screened compounds, apply Fuzzy logic membership functions to calculate membership values for active and inactive compounds to determine the degree to which each compound belongs to the active class or to the inactive class, train an artificial neural network (ANN) to identify active compounds in silico based on the Fuzzy logic membership functions, and process another subset of compounds in silico to identify active and inactive compounds using the trained artificial neural network.

Abstract: Disclosed are techniques for extracting, identifying, and consuming imprecise temporal elements (“ITEs”). A user input may be received from a client device. A prediction may be generated of one or more time intervals to which the user input refers based upon an ITE model. The user input may be associated with the prediction, and provided to the client device.

Abstract: Systems, methods, and computer-readable media are provided for facilitating clinical decision making by directing the emission of computer-generated health-care related recommendations towards contexts in which the recipient will likely find the recommendations salient and will likely welcome them and act upon them. ‘Uptake’ of computer-generated recommendations for diagnostic tests or therapeutic interventions is thereby substantially increased, and ‘alert fatigue’ is substantially decreased. Embodiments of our technology overcome certain drawbacks associated with the prior art by providing a means for ascertaining which decision-support recommendations are likely to be favorably considered by the recipient and acted-upon (recommendation ‘uptake’). System and method embodiments for providing a predicted probability of user uptake of a context-specific system-generated recommendation patient are disclosed herein and for applying that information to decide whether or not to emit the relevant recommendation.

Abstract: Methods, apparatus, systems, and computer-readable media are provided for obtaining user interaction data indicative of interaction by a user with an application executing on a computing device, determining, based on the user interaction data, a likelihood that the user failed to complete a task the user started with the application executing on the computing device, and selectively causing, based on the likelihood, a task-completion reminder to be presented to the user in a manner selected based at least in part on historical reminder consumption.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for decreasing neural network inference times using softmax approximation. One of the methods includes maintaining data specifying a respective softmax weight vector for each output in a vocabulary of possible neural network outputs; receiving a neural network input; processing the neural network input using one or more initial neural network layers to generate a context vector for the neural network input; and generating an approximate score distribution over the vocabulary of possible neural network outputs for the neural network input, comprising: processing the context vector using a screening model configured to predict a proper subset of the vocabulary for the context input; and generating a respective logit for each output that is in the proper subset, comprising applying the softmax weight vector for the output to the context vector.

Abstract: The modeling of an impression effect may include generating a content item impression effect distribution. A classification model may be used to determine a period of the content item impression effect distribution based on one or more accessed impression effect parameters. A value for a content item may be determined based, at least in part, on the determined period and a bid associated with the content item. A content item may be selected based on the determined value and data to display the selected content item may be transmitted. In some instances, the determined period may be used to determine or select predictive model for the determined period that outputs a factor to modify the determined value.

Abstract: Embodiments of a system for determining personal attributes based on public interaction data are illustrated. In one embodiment, the system employs a process for predicting personal attributes based on public interaction data by constructing matrices based on user interactions drawn from public posts on a social media website. The process may further learn a compact representation for a plurality of users based on public posts using the matrices, extract the compact representation of one or more users that have been labeled, and apply a classifier to learn about a particular personal attribute. Through this, a prediction of personal attributes of users that have not been labeled may be obtained.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing an input sequence using a recurrent neural network to generate an output for the input sequence. One of the methods includes receiving the input sequence; generating a doubled sequence comprising a first instance of the input sequence followed by a second instance of the input sequence; and processing the doubled sequence using the recurrent neural network to generate the output for the input sequence.