A computer-implemented method is provided for evaluating a pilot tone signal. In the method, the pilot tone signal is recorded using a high-frequency coil arrangement of a magnetic resonance facility and describes a movement of a patient. The method also includes extracting movement information assigned to a movement component, (e.g., a respiratory movement). A breakdown or decomposition of the pilot tone signal is effected on a basis of signal components having assigned weightings and for the purpose of determining the movement information, a part of a base which is assigned to the movement component is selected by a selection criterion. For the purpose of determining the base and the weightings, a non-negative matrix factorization is performed, in the context of which a signal matrix, which is formed from the pilot tone signal and is in particular non-negative, is formulated as a product of a non-negative signal component matrix that describes the base and a non-negative weighting matrix that describes the weightings.

A computer-implemented method is provided for evaluating a pilot tone signal. In the method, the pilot tone signal is recorded using a high-frequency coil arrangement of a magnetic resonance facility and describes a movement of a patient. The method also includes extracting movement information assigned to a movement component, (e.g., a respiratory movement). A breakdown or decomposition of the pilot tone signal is effected on a basis of signal components having assigned weightings and for the purpose of determining the movement information, a part of a base which is assigned to the movement component is selected by a selection criterion. For the purpose of determining the base and the weightings, a non-negative matrix factorization is performed, in the context of which a signal matrix, which is formed from the pilot tone signal and is in particular non-negative, is formulated as a product of a non-negative signal component matrix that describes the base and a non-negative weighting matrix that describes the weightings.

Machine-learning methods and apparatus are provided to solve blind source separation problems with an unknown number of sources and having a signal propagation model with features such as wave-like propagation, medium-dependent velocity, attenuation, diffusion, and/or advection, between sources and sensors. In exemplary embodiments, multiple trials of non-negative matrix factorization are performed for a fixed number of sources, with selection criteria applied to determine successful trials. A semi-supervised clustering procedure is applied to trial results, and the clustering results are evaluated for robustness using measures for reconstruction quality and cluster separation. The number of sources is determined by comparing these measures for different trial numbers of sources. Source locations and parameters of the signal propagation model can also be determined. Disclosed methods are applicable to a wide range of spatial problems including chemical dispersal, pressure transients, and electromagnetic signals, and also to non-spatial problems such as cancer mutation.

The invention provides a non-negative matrix factorization face recognition method and system based on kernel machine learning, which comprises five steps. The invention has the following beneficial effects: the invention avoids the learning of the inaccurate pre-image matrix by directly learning two kernel matrices, K.sub.wx and K.sub.ww, and avoids the derivation of the kernel function in the iterative formula by changing the learning object, so that there is no limit to the selection of kernel function and a general algorithm for any kernel function is obtained.

The method includes: obtaining a plurality of pieces of feature data; automatically performing two different types of nonlinear combination processing operations on the plurality of pieces of feature data to obtain two groups of processed data, where the two groups of processed data include a group of higher-order data and a group of lower-order data, the higher-order data is related to a nonlinear combination of m pieces of feature data in the plurality of pieces of feature data, and the lower-order data is related to a nonlinear combination of n pieces of feature data in the plurality of pieces of feature data, where m3, and m>n2; and determining prediction data based on a plurality of pieces of target data, where the plurality of pieces of target data include the two groups of processed data.

A method of identifying malicious activity in a computer data sequence includes providing provided the computer data sequence to a network configured to convert the computer data sequence from a high-dimensional space to a low-dimensional space, and processing the computer data sequence in the low-dimensional space to generate an approximately Gaussian distribution. The processed computer data sequence converted to the low dimensional space is evaluated relative to the approximately Gaussian distribution to determine whether the computer data sequence is likely malicious or likely benign, and an output is provided indicating whether the computer data sequence is likely malicious or likely benign.

Described is a system for classifying novel objects in imagery. In operation, the system extracts salient patches from a plurality of unannotated images using a multi-layer network. Activations of the multi-layer network are clustered into key attribute, with the key attributes being displayed to a user on a display, thereby prompting the user to annotate the key attributes with class label. An attribute database is then generated based on user prompted annotations of the key attributes. A test image can then be passed through the system, allowing the system to classify at least one object in the test image by identifying an object class in the attribute database. Finally, a device can be caused to operate or maneuver based on the classification of the at least one object in the test image.

Provided are an apparatus and a method using a convolutional neural network (CNN) including a plurality of convolution layers in the field of artificial intelligence (AI) systems and applications thereof. A computing apparatus using a CNN including a plurality of convolution layers includes a memory storing one or more instructions; and one or more processors configured to execute the one or more instructions stored in the memory to obtain input data; identify a filter for performing a convolution operation with respect to the input data, on one of the plurality of convolution layers; identify a plurality of sub-filters corresponding to different filtering regions within the filter; provide a plurality of feature maps based on the plurality of sub-filters; and obtain output data, based on the plurality of feature maps.

A device may receive activity data identifying activities of a process performed by users via user devices. The device may receive baseline data identifying baselines associated with the process and variant data identifying variants from the baselines. The device may apply a sequence alignment model, to the activity data and the baseline data, to determine first similar sequences associated with the activities and the baselines and may apply the sequence alignment model, to the activity data and the variant data, to determine second similar sequences associated with the activities and the variants. The device may determine, based on the first similar sequences, first label data identifying first labels for the activities and may determine, based on the second similar sequences, second label data identifying second labels for the activities. The device may generate a process diagram based on the activity data, the first label data, and the second label data.

A device may receive activity data identifying activities of a process performed by users via user devices. The device may receive baseline data identifying baselines associated with the process and variant data identifying variants from the baselines. The device may apply a sequence alignment model, to the activity data and the baseline data, to determine first similar sequences associated with the activities and the baselines and may apply the sequence alignment model, to the activity data and the variant data, to determine second similar sequences associated with the activities and the variants. The device may determine, based on the first similar sequences, first label data identifying first labels for the activities and may determine, based on the second similar sequences, second label data identifying second labels for the activities. The device may generate a process diagram based on the activity data, the first label data, and the second label data.