Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating neural network architectures. One of the methods includes receiving a request to determine an architecture for a task neural network; maintaining data specifying a plurality of candidate architectures for the task neural network; repeatedly performing operations comprising: selecting one or more candidate architectures in the maintained data to be modified; generating a new candidate architecture from the selected candidate architecture by, for each hyperparameter in the set of hyperparameters, selecting the value for the hyperparameter for the new candidate architecture; and adding data specifying the new candidate architecture to the maintained data; and selecting, as the final architecture for the task neural network, one of the candidate architectures specified in the maintained data.

Systems and methods are described herein for generating potential feature combinations for a new item. A neural network may be utilized to identify positive and/or negative sentiment phrases from textual data. Each sentiment phrase may correspond to particular features of existing items. A machine-learning model may utilize the sentiment phrases and their corresponding features to generate a set of potential feature combinations for a new item. The potential feature combinations may be scored, for example, based on an amount by which a potential feature combination differs from known feature combinations of existing items. One or more potential feature combinations may be provided in a feature recommendation. Feedback (e.g., human feedback, sales data, page views for similar items, and the like) may be obtained and utilized to retrain the machine-learning model to better identify subsequent feature combinations that may be desirable and/or practical to manufacture.

A system and method for predicting mild cognitive impairment (MCI) related diagnosis and prognosis utilizing hybrid machine learning. More specifically, the system and method produce predictions of MCI conversions to dementia and prognosis related thereof. Using available medical imaging and non-imaging data a diagnosis and prognosis model is trained using transfer learning. A platform may then receive a request from a clinician for a target patient's diagnosis or prognosis. The target patient's medical data is retrieved and used to create a model for the target patient. Then details of the target patient's model and the diagnosis and prognosis model are compared, a prediction is generated, and the prediction is returned to the clinician. As new medical data becomes available it is fed into the respective model to improve accuracy and update predictions.

Example methods, apparatuses, and/or articles of manufacture are disclosed that may be implemented, in whole or in part, using one or more computing devices to select options for decisions in connection with design features of a computing device. In a particular implementation, design options for two or more design decisions of neural network processing device may be selected based, at least in part, on combination of function values that are computed based, at least in part, on a tensor expressing sample neural network weights.

Provided is a method for implementing reinforcement learning by a neural network. The method may include performing, for each epoch of a first predetermined number of epochs, a second predetermined number of training iterations and a third predetermined number of testing iterations using a first neural network. The first neural network may include a first set of parameters, the training iterations may include a first set of hyperparameters, and the testing iterations may include a second set of hyperparameters. The testing iterations may be divided into segments, and each segment may include a fourth predetermined number of testing iterations. A first pattern may be determined based on at least one of the segments. At least one of the first set of hyperparameters or the second set of hyperparameters may be adjusted based on the pattern. A system and computer program product are also disclosed.

Provided is a method for implementing reinforcement learning by a neural network. The method may include performing, for each epoch of a first predetermined number of epochs, a second predetermined number of training iterations and a third predetermined number of testing iterations using a first neural network. The first neural network may include a first set of parameters, the training iterations may include a first set of hyperparameters, and the testing iterations may include a second set of hyperparameters. The testing iterations may be divided into segments, and each segment may include a fourth predetermined number of testing iterations. A first pattern may be determined based on at least one of the segments. At least one of the first set of hyperparameters or the second set of hyperparameters may be adjusted based on the pattern. A system and computer program product are also disclosed.

According to a first aspect of the present disclosure, a method for facilitating the detection of one or more time series patterns is conceived, comprising building one or more artificial neural networks, wherein, for at least one time series pattern to be detected, a specific one of said artificial neural networks is built. According to a second aspect of the present disclosure, a corresponding computer program is provided. According to a third aspect of the present disclosure, a non-transitory computer-readable medium is provided that comprises a computer program of the kind set forth. According to a fourth aspect of the present disclosure, a corresponding system for facilitating the detection of one or more time series patterns is provided.

An efficient technique of machine learning is provided for training a plurality of convolutional neural networks (CNNs) with increased speed and accuracy using a genetic evolutionary model. A plurality of artificial chromosomes may be stored representing weights of artificial neuron connections of the plurality of respective CNNs. A plurality of pairs of the chromosomes may be recombined to generate, for each pair, a new chromosome (with a different set of weights than in either chromosome of the pair) by selecting entire filters as inseparable groups of a plurality of weights from each of the pair of chromosomes (e.g., “filter-by-filter” recombination). A plurality of weights of each of the new or original plurality of chromosomes may be mutated by propagating recursive error corrections incrementally throughout the CNN. A small random sampling of weights may optionally be further mutated to zero, random values, or a sum of current and random values.

An efficient technique of machine learning is provided for training a plurality of convolutional neural networks (CNNs) with increased speed and accuracy using a genetic evolutionary model. A plurality of artificial chromosomes may be stored representing weights of artificial neuron connections of the plurality of respective CNNs. A plurality of pairs of the chromosomes may be recombined to generate, for each pair, a new chromosome (with a different set of weights than in either chromosome of the pair) by selecting entire filters as inseparable groups of a plurality of weights from each of the pair of chromosomes (e.g., “filter-by-filter” recombination). A plurality of weights of each of the new or original plurality of chromosomes may be mutated by propagating recursive error corrections incrementally throughout the CNN. A small random sampling of weights may optionally be further mutated to zero, random values, or a sum of current and random values.

Provided is an LSTM-based hot-rolling roll-bending force predicting method including the steps of acquiring final rolling data of a stand of a stainless steel rolling mill when performing a hot rolling process, and dividing the data into a training set traindata and a test set testdata; normalizing the traindata; building a matrix P; using a last row of the matrix P as a label of the training set, namely a true value; calculating and updating an output value and the true value of a network; after network training is completed, taking the last m output data of the LSTM network as an input at a next moment, and then obtaining an output of the network at the next moment, wherein the output is a predicted value of the roll-bending force at the next moment; repeating the steps until a sufficient number of prediction data is obtained; and comparing the processed data with the true value in the testdata to check the validity of the network.