Abstract: Systems, methods, and software for developing application usage information are provided herein. An exemplary method includes identifying data sources that relate to application usage activity of one or more users, determining activity signals related to at least data analysis by the one or more users from among the application usage activity, and applying the activity signals to a knowledge graphing service that infers data analysis preferences from among the activity signals. The method also includes providing ones of the data analysis preferences for use by data insight services that establish data insight objects visualizing target datasets based at least on the ones of the data analysis preferences.

Abstract: One or more implementations of the present specification provide a risk identification model building method, and a risk identification method. From data of a target user, user state records of the target user within a predetermined duration of time are extracted, wherein the user state records include a plurality of user operations and/or a plurality of system events. The user state records are sorted based on corresponding occurrence times. A user state sequence is generated based on sorted user state records. The generated user state sequence is converted into a sequence feature. A risk identification result is generated based on a previously trained risk identification model that takes as input the sequence feature generated from the user state sequence.

Abstract: Various technologies pertaining to allocating computing resources of a neuromorphic computing system are described herein. Subgraphs of a neural algorithm graph to be executed by the neuromorphic computing system are identified. The subgraphs are each executed by a group of neuron circuits serially. Output data generated by execution of the subgraphs are provided to the same or a second group of neuron circuits at a same time or with associated timing data indicative of a time at which the output data was generated. The same or second group of neuron circuits performs one or more processing operations based upon the output data.

Abstract: A heuristic engine includes capabilities to collect an unstructured data set and a current business context. Providing a heuristic algorithm, executing within the engine, with the data set and the context may allow determination of predicted future contexts and subsequent actions that refine and improve the quality of service provided to a customer. Such heuristic algorithms may learn from past data transactions and appropriate correlations with events and available data.

Abstract: Graphical interactive prediction evaluation is provided. An extrapolation threshold value is computed using an extrapolation threshold function with an explanatory variable value of each of a plurality of explanatory variables read for each observation vector of a plurality of observation vectors. A model is fit to the observation vectors. Model results are presented in a display that include a first value for each explanatory variable. An indicator of a second value of at least one of the explanatory variables that is different from the first value is received. An extrapolation value is computed using an extrapolation function with the second value and the first value of others of the explanatory variables. The extrapolation value is compared to the extrapolation threshold value. An extrapolation indicator is presented in the display when the comparison indicates that the second value is an extrapolation.

Abstract: In a method of operating a neural network device, a plurality of consecutive input data is received by an input layer. A delta data is generated by the input layer based on a difference between a current input data and a previous input data. A first current feature is generated by a first linear layer based on a first delta feature generated by performing a first linear operation on the delta data and a first previous feature. A second delta feature is generated by a first nonlinear layer based on a second current feature generated by performing a first nonlinear operation on the first current feature and a second previous feature. A third current feature is generated by a second linear layer based on a third delta feature generated by performing a second linear operation on the second delta feature and a third previous feature.

Abstract: A long-term memory network method and system for text comprehension. A recurrent neural network can be provided, which includes an external memory module and a long-short term memory unit, wherein said recurrent neural network encodes raw text information into vector representations, forms memories, finds relevant sentences to answer questions, and generates multi-word answers to said questions utilizing the long short term memory unit.

Abstract: A framework for generating optimized recommendations is described herein. For example, an optimized customer recommendation engine is described herein. Customer data is collected and pre-processed into a data model. Recommendations are calculated and provided by an aggregated method. The aggregated output is generated based on the outputs of a real-time prediction model and an offline modeling process. The real-time prediction model may be an online modeling training technique based on support vector machines (SVM) to classify customers and provide quick recommendations. The offline modeling process may be a learning process based on a back-propagation artificial neural network (BP-ANN) to provide with reliable predictions. Validation may be introduced to evaluate the accuracy of the recommendation model.

Abstract: Described is an apparatus which comprises: a first clocking source having a first divider; a second clocking source having a second divider, wherein the first and second clocking sources are inductively coupled; and calibration logic to monitor clock signals associated with the first and second clocking sources and to generate at least one calibration code for adjusting at least one divider ratio of the first or second dividers according to the monitored clock signals.

Abstract: Technologies are described herein for injecting elements into artificial intelligence content. According to some examples, content generated from an artificial intelligence source is received, facts are determined from the content, and terms are selected for use based on the facts. The terms are used to modify or are added to the content to generate modified artificial intelligence content.

Abstract: A load control device may be configured to control multiple characteristics of one or more electrical loads such as the intensity and color of a lighting load. The load control device may include concentric rotating portions for adjusting the multiple characteristics. A control circuit of the load control device may be configured to generate control data for controlling one or more of the characteristics of the electrical loads in response to rotations of the concentric rotating portions. The control circuit may be further configured to provide feedback regarding the control being applied on one or more visual indicators. The load control device may be a wall-mounted dimmer device or a battery-powered remote control device.

Abstract: Disclosed is a diagnostic method and system including the processing of historical event logs generated by one or more devices. According to an exemplary embodiment, a diagnostic system includes an event log acquisition module, an event classification module classifying event logs acquired, and a diagnostic module generating a labeled version of the historical event log including labels provided by the classification module. The event classification module is trained using supervised machine learning techniques.

Abstract: Systems and methods for predicting the outcome of a business entity are presented. In embodiments, a system may receive explicit data reporting or indicating activities of a business entity, and other data from which information regarding the activities or level of operations of the entity may be inferred. Using one or more data processors, the system may generate inferred data regarding the business entity from a selected portion of the other data, and use at least some of the explicit data and the inferred data to determine which one of a series of defined sequential active states of development the entity currently is in. The system may further, using the result of the determination as the current state of the business, predict a final stage of the business entity, and a probability of evolving to that final stage from the current state. Other embodiments may be disclosed or claimed.

Abstract: An apparatus to facilitate optimization of a neural network (NN) is disclosed. The apparatus includes optimization logic to define a NN topology having one or more macro layers, adjust the one or more macro layers to adapt to input and output components of the NN and train the NN based on the one or more macro layers.

Abstract: The disclosure provides an approach for learning to schedule control fragments for physics-based virtual character simulations and physical robot control. Given precomputed tracking controllers, a simulation application segments the controllers into control fragments and learns a scheduler that selects control fragments at runtime to accomplish a task. In one embodiment, each scheduler may be modeled with a Q-network that maps a high-level representation of the state of the simulation to a control fragment for execution. In such a case, the deep Q-learning algorithm applied to learn the Q-network schedulers may be adapted to use a reward function that prefers the original controller sequence and an exploration strategy that gives more chance to in-sequence control fragments than to out-of-sequence control fragments.

Abstract: A chatbot in the context of a chat group messaging is described. The chat group can include a plurality of users and a chatbot. A set of rules can be defined for the users of the group granting each user a privilege status. The chatbot can receive a request through a message transmitted to the chat group. The chatbot can discern a task associated with the message, and perform the task or ask another module to perform the task. Once the task is performed, the chatbot can report the results to the chat group. The chatbot can include a conflict resolution module which can resolve conflicts. The conflict resolution module can use each user's privilege status to resolve the conflicts.

Abstract: A framework for improving user interfaces, and in particular for improving user interfaces for displaying and interacting with predictive analytics, is described herein. In one embodiment, a user interface template renders predictive models and enables visually interacting with data to discover hidden insights and relationships in the data. The user interface template determines, based on the metadata and data annotations, how to display the supplied data. By encapsulating complex code necessary to render predictive models and enable visually interacting with data, the amount of frontend code required to implement predictive analytic functionality is reduced, defect rates are reduced, while design consistency is improved.

Abstract: Methods, systems, and computer programs are presented for updating the data model of a structural damage predictor after an earthquake. One method includes an operation for identifying features for a structure and fragility functions for predicting structural damage to the structure, the fragility functions being stored in a database. The method further includes an operation for estimating a first damage to the structure after an earthquake utilizing a damage-estimation algorithm and the fragility functions for the structure. One or more of the fragility functions are changed based on the first damage to the structure when the first damage to the structure is above a predetermined damage threshold. The method further includes operations for accessing shaking data for a new earthquake, and for estimating a second damage to the structure after the new earthquake utilizing the damage-estimation algorithm and the fragility functions for the structure.

Abstract: Methods, systems, and computer programs are presented for updating estimates of damage caused by a disaster based on newly acquired damage data. One method includes operations for generating block damage estimates in a geographical region after an event (e.g., a natural disaster such as an earthquake or a tornado), accessing input damage data for one or more buildings within a first block, and adjusting the block damage estimate of the first block based on the input damage data. One or more related blocks within a threshold distance from the first block are identified, and for each related block, a respective propagation coefficient is determined based on a comparison of features of the first block with features of each related block. The block damage estimate for the one or more related blocks is recalculated based on the respective propagation coefficient.

Abstract: A neural network-based classifier system can receive a query including a media signal and, in response, provide an indication that a particular received query corresponds to a known media type or media class. The neural network-based classifier system can select and apply various models to facilitate media classification. In an example embodiment, classifying a media query includes accessing digital media data and a context parameter from a first device. A model for use with the network-based classifier system can be selected based on the context parameter. In an example embodiment, the network-based classifier system provides a media type probability index for the digital media data using the selected model and spectral features corresponding to the digital media data. In an example embodiment, the digital media data includes an audio or video signal sample.