Abstract: The present disclosure provides methods for applying artificial neural networks to flow cytometry data generated from biological samples to diagnose and characterize cancer in a subject. The disclosure also provides methods of training, testing, and validating artificial neural networks.

Abstract: An apparatus and method for correlating events in a smart home system as a pattern. The apparatus and method include collecting from smart home devices, state change events of the smart home system, determining whether a series of the collected state change events are a known pattern, requesting, when the series of the collected state change events is an unknown pattern, users of the smart home system to identify what caused the collected state change events, and judging, by the smart home users, a best reason among the identified causes of the collected state change events.

Abstract: During a training phase of a machine learning model, representations of at least some nodes of a decision tree are generated and stored on persistent storage in depth-first order. A respective predictive utility metric (PUM) value is determined for one or more nodes, indicating expected contributions of the nodes to a prediction of the model. A particular node is selected for removal from the tree based at least partly on its PUM value. A modified version of the tree, with the particular node removed, is stored for obtaining a prediction.

Abstract: Described herein are systems and methods for correcting a data set and classifying the data set in an integrated manner. A training data set, a training class set, and a test data set are received. A first classifier is generated for the training data set by applying a machine learning technique to the training data set and the training class set, and a first test class set is generated by classifying the elements in the test data set according to the first classifier. For each of multiple iterations, the training data set is transformed, the test data set is transformed, and a second classifier is generated by applying a machine learning technique to the transformed training data set. A second test class set is generated according to the second classifier, and the first test class set is compared to the second test class set.

Abstract: Systems and methods that quantify document relevance for a document relative to a training corpus and select a best match or best matches are provided herein. Methods may include generating an example-based explanation for relevancy of a document to a training corpus by executing a support vector machine classifier, the support vector machine classifier performing a centroid classification of a relevant document in a term frequency-inverse document frequency features space relative to training examples in a training corpus, and generating an example-based explanation by selecting a best match for the relevant document from the training examples based upon the centroid classification.

Abstract: Systems and methods for managing content delivery functionalities based on machine learning models are provided. In one aspect, content requests are routed in accordance with clusters of historical content requests to optimize cache performance. In another aspect, content delivery strategies for responding to content requests are determined based on a model trained on data related to historical content requests. The model may also be used to determine above-the-fold configurations for rendering responses to content requests. In some embodiments, portions of the model can be executed on client computing devices.

Abstract: Systems and methods for managing content delivery functionalities based on machine learning models are provided. In one aspect, content requests are routed in accordance with clusters of historical content requests to optimize cache performance. In another aspect, content delivery strategies for responding to content requests are determined based on a model trained on data related to historical content requests. The model may also be used to determine above-the-fold configurations for rendering responses to content requests. In some embodiments, portions of the model can be executed on client computing devices.

Abstract: Hyper-parameters are selected for training a deep convolutional network by selecting a number of network architectures as part of a database. Each of the network architectures includes one or more local logistic regression layer and is trained to generate a corresponding validation error that is stored in the database. A threshold error for identifying a good set of network architectures and a bad set of network architectures may be estimated based on validation errors in the database. The method also includes choosing a next potential hyper-parameter, corresponding to a next network architecture, based on a metric that is a function of the good set of network architectures. The method further includes selecting a network architecture, from among next network architectures, with a lowest validation error.

Abstract: Embodiments are directed towards a machine learning repository for managing machine learning (ML) model envelopes, ML models, model objects, or the like. Questions and model objects may be received by a ML model answer engine. Machine learning (ML) model envelopes may be received based on the questions. The model objects may be compared to parameter models associated with the ML model envelopes. ML model envelopes may be selected based on the comparison such that the model objects satisfy the parameter models of each of the selected ML model envelopes. ML models included in each selected ML model envelope may be executed to provide score values for the model objects and the score values may be included in a report.

Abstract: An apparatus and methods for training and/or operating a robotic device to perform a target task autonomously. The target task execution may be configured based on analysis of sensory context by the robot. Target action may comprise execution of two or more mutually exclusive actions for a given context. The robotic device may be operable in accordance with a persistent switching process. For a given sensor input, the switching process may be trained to select one of two or more alternative actions based on a prior action being executed. Switching process operation may comprise assigning priorities to the available tasks based on the sensory context; the task priorities may be modified during training based on input from a trainer. The predicted task priorities may be filtered by a “persistent winner-take-all process configured to switch from a current task to another task based on the priority breaching a switching threshold.

Abstract: A system (10) for receiving measurement data wherein there is a processing means of the measurement data to create or update at least two models (28) where each model is a representation of a real-world object, event or phenomenon. There is a means for processing the characteristics (32) of the models to establish relationships (34) between the models and to generate alarms (40) when the relationships meet a certain criteria.

Abstract: A neural processing engine may perform processing within a neural processing system and/or artificial neural network. The neural processing engine may be configured to effectively and efficiently perform the type of processing required in implementing a neural processing system and/or an artificial neural network. This configuration may facilitate such processing with neural processing engines having an enhanced computational density and/or processor density with respect to conventional processing units.

Abstract: An approach is provided in which a QA system ingests traditional sources, which includes traditional terms, into a domain dictionary. Next, the QA system ingests crowd-based sources that include crowd-based terms and corresponding crowd-based metadata. In turn, the QA system calculates weightings pertaining to the traditional terms based upon the crowd-based metadata. When the QA system receives a question from a requestor that includes question terms, the QA system identifies an answer to the question based on the calculated weightings pertaining to the traditional terms that are relevant to the question terms.