1206 results about "Generative model" patented technology

The present invention relates to a scaleable automatic method of using multiple techniques to generate models and combinations of models from data and prior knowledge. The system provides unprecedented ease of use in that many of the choices of technique and parameters are explored automatically by the system, without burdening the user, and provides scaleable learning over distributed processors to achieve speed and data-handling capacity to satisfy the most demanding requirements.

A device for modeling an integrated enterprise includes a first tool for constructing a model of the integrated enterprise and a second tool for analyzing calls, contained in the constructed model, between applications of the integrated enterprise and a database shared thereby. The call analysis tool identifies data attribute imbalances in calls between the applications of the integrated enterprise and the shared database. Data attribute imbalances result if the calls attempt to consume data which was never produced or if the calls produce data which is never consumed. The call analysis tool also generates producer / consumer maps of the data attributes used in the model, producer exception reports which identify data attributes which are consumed but never produced and consumer exception reports which identify data attributes which are produced but never consumed.

An automated image processing system and method are provided for class-based segmentation of a digital image. The method includes extracting a plurality of patches of an input image. For each patch, at least one feature is extracted. The feature may be a high level feature which is derived from the application of a generative model to a representation of low level feature(s) of the patch. For each patch, and for at least one object class from a set of object classes, a relevance score for the patch, based on the at least one feature, is computed. For at least some or all of the pixels of the image, a relevance score for the at least one object class based on the patch scores is computed. An object class is assigned to each of the pixels based on the computed relevance score for the at least one object class, allowing the image to be segmented and the segments labeled, based on object class.

A system generates a model based on feature data relating to different features of a link from a linking document to a linked document and user behavior data relating to navigational actions associated with the link. The system also assigns a rank to a document based on the model.

An “Interactive Word Lattice” provides a user interface for interacting with and selecting user-modifiable paths through a lattice-based representation of alternative suggested text segments in response to a user's text segment input, such as phrases, sentences, paragraphs, entire documents, etc. More specifically, the user input is provided to a trained paraphrase generation model that returns a plurality of alternative text segments having the same or similar meaning as the original user input. An interactive graphical lattice-based representation of the alternative text segments is then presented to the user. One or more words of each alternative text segment represents a “node” of the lattice, while each connection between nodes represents a lattice “edge. Both nodes and edges are user modifiable. Each possible path through the lattice corresponds to a different alternative text segment. Users select a path through the lattice to select an alternative text to the original input.

A device capable of achieving recognition at a high accuracy and with fewer calculations and which utilizes an HMM. The present device has a vector quantizing circuit generating a model by quantizing vectors of a training pattern having a vector series, and converting the vectors into a label series of clusters to which they belong, a continuous distribution probability density HMM generating circuit for generating a continuous distribution probability density HMM from a quantized vector series corresponding to each label of the label series, and a label incidence calculating circuit for calculating the incidence of the labels in each state from the training vectors classified in the same clusters and the continuous distribution probability density HMM.

Methods and systems for generating simulated images from design information are provided. One system includes one or more computer subsystems and one or more components executed by the computer subsystem(s), which include a generative model. The generative model includes two or more encoder layers configured for determining features of design information for a specimen. The generative model also includes two or more decoder layers configured for generating one or more simulated images from the determined features. The simulated image(s) illustrate how the design information formed on the specimen appears in one or more actual images of the specimen generated by an imaging system.

A system for generating a model is provided. The system generates, or selects, candidate conditions and generates, or otherwise obtains, statistics regarding the candidate conditions. The system also forms rules based, at least in part, on the statistics and the candidate conditions and selectively adds the rules to the model.

The invention provides a method and apparatus for the dynamic generation and regeneration of software objects that can have logic that specifies structure, functionality and behavior in a computing system. One embodiment is the regeneration function that produces related, but different instances of software objects, including System Logic Objects. The regeneration function comprises ordered sequences of tasks, whose definition, and parameterization are obtained from a Generative Model, and which are performed by software objects called Builders. Each Builder accepts a set of input parameters and a reference to a container that contains zero or more generated software objects, including System Logic Objects. The regeneration function directs each Builder to work on the contents of the container, add, modify, and delete objects, and return the container with updated contents. The behavior of the regeneration function can be altered by disabling selected Builders, and by changing the inputs to the Builders.

A control system comprises a controller for positioning an actuator in a working fluid flow and a model processor for directing the controller as a function of a model feedback. The model processor comprises an output module, a comparator and an estimator. The output module generates the model feedback as a function of a constraint, a model state and a model input describing fluid parameters measured along the working fluid flow. The comparator generates an error by comparing the model feedback to the model input. The estimator generates the constraint and the model state, such that the error is minimized.

A method, computer readable medium and system are provided which generate a model driven form user interface to represent an application model. The method includes selecting which of multiple different logical form types to use to generate the form user interface to represent the application model. The method also includes providing a first map. A display target independent logical form is generated using the application model, the selected form type and the first map.

In various embodiments, the systems and methods described herein relate to generative models. The generative models may be trained using machine learning approaches, with training sets comprising chemical compounds and biological or chemical information that relate to the chemical compounds. Deep learning architectures may be used. In various embodiments, the generative models are used to generate chemical compounds that have desired characteristics, e.g. activity against a selected target. The generative models may be used to generate chemical compounds that satisfy multiple requirements.

Methods and systems for detecting defects on a specimen are provided. One system includes a generative model. The generative model includes a non-linear network configured for mapping blocks of pixels of an input feature map volume into labels. The labels are indicative of one or more defect-related characteristics of the blocks. The system inputs a single test image into the generative model, which determines features of blocks of pixels in the single test image and determines labels for the blocks based on the mapping. The system detects defects on the specimen based on the determined labels.

System and method for managing batch fermentation in biofuel production. An optimizer executes a nonlinear multivariate predictive model of a batch fermentation process in accordance with an end of batch objective specifying a target end of batch biofuel concentration to determine an optimal batch trajectory over a temporal control horizon specifying a biofuel and / or sugar concentration trajectory over the batch fermentation process. A nonlinear control model for the batch fermentation process that includes the temporal control horizon driven by biofuel concentration during the batch fermentation process is executed per the determined optimal batch trajectory using received process information as input, thereby generating model output including target values for manipulated variables for the batch fermentation process, including batch fermentation temperature. The batch fermentation process is controlled per the target values to produce biofuel in accordance with the determined optimal batch trajectory, to substantially optimize the end of batch biofuel yield.

Methods and systems are provided for a systematic approach to computer aided inventing. In a modeling environment, a model representing any item, composition or process can be defined and configured using the lexicon and specification of an innovation database. In the model, objects can be identified, defined, and configured to provide the model with constituent products, components, features and materials. An innovation engine automatically applies one or more morphs from the innovation database to the objects of the model to generate morphed versions of the model that may provide innovations of or inventions to the item, composition or process represented by the model.

In an image classification method, a feature vector representing an input image is generated by unsupervised operations including extracting local descriptors from patches distributed over the input image, and a classification value for the input image is generated by applying a neural network (NN) to the feature vector. Extracting the feature vector may include encoding the local descriptors extracted from each patch using a generative model, such as Fisher vector encoding, aggregating the encoded local descriptors to form a vector, projecting the vector into a space of lower dimensionality, for example using Principal Component Analysis (PCA), and normalizing the feature vector of lower dimensionality to produce the feature vector representing the input image. A set of mid-level features representing the input image may be generated as the output of an intermediate layer of the NN.

The invention discloses a deep neural network and reinforcement learning-based generative machine reading comprehension method. According to the method, texts and questions are encoded through an attention mechanism-combined deep neural network so as to form question information-fused text vector expressions, and decoding is carried out through a unidirectional LSTM decoder so as to gradually generate corresponding answer texts. According to the reading comprehension method, the advantages of extractive models and generative models are fused, training is carried out by adoption of a multi-taskcombined optimization manner, and a reinforcement learning method is used in the training process, so that benefit is brought to generate more correct and fluent answer texts.

A system for developing a computer implemented service, for deployment on computing infrastructure, generates variants of the model by automatically choosing values for a limited set of design variables, and evaluates the variants in operation. A model manager (187) stores in a model repository (107) a current variant (57) and at least some previous variants, and their evaluation results and derivation trails, the generating part being arranged to use the evaluation results and the derivation trails to generate a next current variant. Such use of the repository can help make more efficient the search by the model manager for variants that work well. In particular the derivation trails and evaluations can make it easier to determine when to revert to a preceding variant or determine what new design choices to try next.

A method for quantifying cardiac desynchrony of the right and left ventricles includes obtaining cardiac acquisition data from a medical imaging system, and determining a movement profile from the cardiac acquisition data. The movement profile is directed toward identifying at least one of a time-based contraction parameter for a region of the left ventricle (LV), and a displacement-based contraction parameter for a region of the LV. The determined movement profile is visually displayed by generating a 3D model therefrom.

A control system comprises a controller for positioning an actuator in a working fluid flow and a model processor for directing the controller as a function of a model feedback. The model processor comprises an output module, a comparator and an estimator. The output module generates the model feedback as a function of a constraint, a model state and a model input describing fluid parameters measured along the working fluid flow. The comparator generates an error by comparing the model feedback to the model input. The estimator generates the constraint and the model state, such that the error is minimized.

Disclosed are apparatus and methods for detecting whether a video is adult or non-adult. In certain embodiments, a learning system is operable to generate one or more models for adult video detection. The model is generated based on a large set of known videos that have been defined as adult or non-adult. Adult detection is then based on this adult detection model. This adult detection model may be applied to selected key frames of an unknown video. In certain implementations, these key frames can be selected from the frames of the unknown video. Each key frame may generally correspond to a frame that contains key portions that are likely relevant for detecting pornographic or adult aspects of the unknown video. By way of examples, key frames may include moving objects, skin, people, etc. In alternative embodiments, a video is not divided into key frames and all frames are analyzed by a learning system to generate a model, as well as by an adult detection system based on such model.

Applications of machine learning techniques such as Latent Dirichlet Allocation (LDA) and author-topic models (ATM) to the problems of mining of user roles to specify access control policies from entitlement as well as logs which contain record of the usage of these entitlements are provided. In one aspect, a method for performing role mining given a plurality of users and a plurality of permissions is provided. The method includes the following steps. At least one generative machine learning technique, e.g., LDA, is used to obtain a probability distribution θ for user-to-role assignments and a probability distribution β for role-to-permission assignments. The probability distribution θ for user-to-role assignments and the probability distribution β for role-to-permission assignments are used to produce a final set of roles, including user-to-role assignments and role-to-permission assignments.

A system, software module, and computer program product for performing clustering based data mining that improved performance in model building, good integration with the various databases throughout the enterprise, flexible specification and adjustment of the models being built, and flexible model arrangement and export capability. The software module for performing clustering based data mining in an electronic data processing system comprises: a model setup block operable to receive client input including information specifying a setup of a clustering data mining models, generate the model setup, and generate parameters for the model setup based on the received information, a modeling algorithms block operable to select and initialize a clustering modeling algorithm based on the generated model setup, a model building block operable to receive training data and build a clustering model using the training data and the selected clustering modeling algorithm and a model scoring block operable to receive scoring data and generate predictions and / or recommendations using the scoring data and the clustering model.

A model for analyzing a cashflow sensitive instrument is described that uses an optimization model of a data set associated with a cashflow sensitive instrument, which optimization model is based at least in part on an interest rate model and a cash-flow model. The interest rate model is at least partially based on at least one random variable used to simulate an underlying distribution on at least one interest rate. A model output is then generated based on the optimization model. The model outputs all optimal cashflow solution for the cashflow sensitive instrument(s) that at least partially optimizes factors of risk and / or cost. Related methods, programs and systems are also provided.

The claimed subject matter provides a system and / or a method that facilitates generating a model from a 3-dimensional (3D) object assembled from 2-dimensional (2D) content. A content aggregator can construct a 3D object from a collection of two or more 2D images each depicting a real entity in a physical real world, wherein the 3D object is constructed by combining the two or more 2D images based upon a respective image perspective. A 3D virtual environment can allow exploration of the 3D object. A model component can extrapolate a true 3D geometric model from the 3D object, wherein the true 3D geometric model is generated to include scaling in proportion to a size within the physical real world.

A method for controlling a gas turbine engine includes: generating model parameter data as a function of prediction error data, which model parameter data includes at least one model parameter that accounts for off-nominal operation of the engine; at least partially compensating an on-board model for the prediction error data using the model parameter data; generating model term data using the on-board model, wherein the on-board model includes at least one model term that accounts for the off-nominal operation of the engine; respectively updating one or more model parameters and one or more model terms of a model-based control algorithm with the model parameter data and model term data; and generating one or more effector signals using the model-based control algorithm.

A computer implemented method comprises analyzing a plurality of attributes of a sample of online documents using a boosted decision tree and generating a model from it. The model is used to predict a click through rate (CTR) of an additional online document based on the analyzing. The predicted CTR is output to a display device, storage medium or network.

System and methods for computer animations of 3D models of heads generated from images of faces is disclosed. A 2D captured image that includes an image of a face can be received and used to generate a static 3D model of a head. A rig can be fit to the static 3D model to generate an animation-ready 3D generative model. Sets of rigs can be parameters that each map to particular sounds or particular facial movement observed in a video. These mappings can be used to generate a playlists of sets of rig parameters based upon received audio or video content. The playlist may be played in synchronization with an audio rendition of the audio content. Methods can receive a captured image, identify taxonomy attributes from the captured image, select a template model for the captured image, and perform a shape solve for the selected template model based on the identified taxonomy attributes.

The invention discloses a machine learning model generation method and a machine learning model generation device. According to a specific embodiment, the method comprises the following steps: generating model parameter combinations, and generating machine learning models respectively corresponding to the model parameter combinations; dividing preset machine learning data into training data and verification data; training the machine learning models in parallel based on the training data; verifying the learning accuracy of the trained machine learning models based on the verification data to get verification scores, and determining an optimal model parameter combination corresponding to a machine learning model to be generated based on the verification scores, and generating a machine learning model corresponding to the optimal model parameter combination. Through the method and the device, machine learning models corresponding to all model parameter combinations can be trained and verified in parallel, the whole process of parameter optimization is improved, and an ideal machine learning model can be generated quickly.