Abstract: The present disclosure provides methods and compositions for controlling asphaltenes in a subterranean formation. The compositions may include a halloysite nanotube. The nanotube has a lumen and a polymer, along with a charged surfactant, may be disposed in the lumen. A wax may be disposed on the nanotube. The wax may be disposed at either or both ends of the nanotube, thereby temporarily preventing the polymer and charged surfactant from leaving the lumen.

Abstract: A method of properly aligning a sliding side door on a vehicle includes first preparing a prototype vehicle and determining how far out of alignment the door is to the body of the prototype vehicle. This amount of misalignment is then used to adjust the attachment position of an anchor plate to a sliding side door of the vehicle, so that the anchor plate is attached at a desired position to the production vehicle, which causes there to be no misalignment between the sliding side door and the body of the production vehicle. A jig is used to hold the anchor plate at the desired location while welding the anchor plate to the door. A roller assembly is then attached to the anchor plate and to a rail on a body of the vehicle.

Abstract: A predictive maintenance server receives data from sensors of equipment. The server uses one or more machine learning models to assign an anomaly score. Responsive to the anomaly score exceeding a threshold value, the server may issue an alert. The machine learning model may be supervised or unsupervised. In one embodiment, the machine learning model use several sensor channels to predict the values of one or more vitals of the equipment and compare the predicted values to the actual measured values of the vitals. The server may assign an anomaly score based on the differences between the predicted values and the measured values. In one embodiment, the machine learning model may be an autoencoder that generates a distribution of the measurement values to determine the likelihood of observing the actual measured values in a normal operation. In one embodiment, the server may use a histogram approach to predict anomaly.

Abstract: Sequence data, such as time series data is analyzed using neural networks, for example, recurrent neural networks. The sequence data is obtained from a source. For example, a sequence data may represent time series data obtained from a sensor. As another example, the sequence of data may represent a sequence of user interactions performed by a user with an online system. The sequences of data are provided as input to a neural network. A feature vector representation of each input sequence data is extracted from the neural network. The feature vector representation is used for clustering the sequence data. Salient features of clusters of sequence data are determined. The salient features of clusters of sequence data are provided for display via a user interface.

Abstract: Training datasets are determined for training neural networks. An input dataset comprising a plurality of samples is provided as training dataset to the neural network. Vector representations of samples of the input dataset are obtained from a hidden layer of the neural network. The samples are clustered using the vector representation. The samples are scored based on a metric that indicates the similarity of the sample to its cluster. A subset of samples is determined by excluding samples that have high similarity with their clusters. The subset of samples is labelled and used for training the neural network.

Abstract: Training datasets are determined for training neural networks. An input dataset comprising a plurality of samples is provided as training dataset to the neural network. Vector representations of samples of the input dataset are obtained from a hidden layer of the neural network. The samples are clustered using the vector representation. The samples are scored based on a metric that indicates the similarity of the sample to its cluster. A subset of samples is determined by excluding samples that have high similarity with their clusters. The subset of samples is labelled and used for training the neural network.

Abstract: A predictive maintenance server receives data from sensors of equipment. The server uses one or more machine learning models to assign an anomaly score. Responsive to the anomaly score exceeding a threshold value, the server may issue an alert. The machine learning model may be supervised or unsupervised. In one embodiment, the machine learning model use several sensor channels to predict the values of one or more vitals of the equipment and compare the predicted values to the actual measured values of the vitals. The server may assign an anomaly score based on the differences between the predicted values and the measured values. In one embodiment, the machine learning model may be an autoencoder that generates a distribution of the measurement values to determine the likelihood of observing the actual measured values in a normal operation. In one embodiment, the server may use a histogram approach to predict anomaly.

Abstract: Embodiments can provide a computer-implemented method for free breathing three dimensional diffusion imaging, the method comprising initiating, via a k-space component processor, diffusion/T2 preparation, comprising generating diffusion contrast; and adjusting one or more of amplitude, duration, and polarity to set a first order moment; applying, via an image data processor, a stack of stars k-space ordering, comprising acquiring a radial/spiral view for all members of a plurality of partitions in a partition-encoding direction; increasing an azimuthal angle until a complete set of radial/spiral views are sampled; and applying diffusion gradients along each of three axis simultaneously; and calculating, via the image data processor, an apparent diffusion coefficient map.

Abstract: A data analysis system stores in-memory representation of a distributed data structure across a plurality of processors of a parallel or distributed system. Client applications interact with the in-memory distributed data structure to process queries using the in-memory distributed data structure and to modify the in-memory distributed data structure. The data analysis system creates uniform resource identifier (URI) to identify each in-memory distributed data structure. The URI can be communicated from one client application to another application using communication mechanisms outside the data analysis system, for example, by email, thereby allowing other client devices to interact with a particular in-memory distributed data structure. The in-memory distributed data structure can be a machine learning model that is trained by one client device and executed by another client device. A client application can interact with the in-memory distributed data structure using different programming languages.

Abstract: Embodiments can provide a computer-implemented method for free breathing three dimensional diffusion imaging, the method comprising initiating, via a k-space component processor, diffusion/T2 preparation, comprising generating diffusion contrast; and adjusting one or more of amplitude, duration, and polarity to set a first order moment; applying, via an image data processor, a stack of stars k-space ordering, comprising acquiring a radial/spiral view for all members of a plurality of partitions in a partition-encoding direction; increasing an azimuthal angle until a complete set of radial/spiral views are sampled; and applying diffusion gradients along each of three axis simultaneously; and calculating, via the image data processor, an apparent diffusion coefficient map.

Abstract: A data analysis system stores in-memory representation of a distributed data structure across a plurality of processors of a parallel or distributed system. Client applications interact with the in-memory distributed data structure to process queries using the in-memory distributed data structure and to modify the in-memory distributed data structure. The data analysis system creates uniform resource identifier (URI) to identify each in-memory distributed data structure. The URI can be communicated from one client application to another application using communication mechanisms outside the data analysis system, for example, by email, thereby allowing other client devices to interact with a particular in-memory distributed data structure. The in-memory distributed data structure can be a machine learning model that is trained by one client device and executed by another client device. A client application can interact with the in-memory distributed data structure using different programming languages.

Abstract: A data analysis system stores in-memory representation of a distributed data structure across a plurality of processors of a parallel or distributed system. Client applications interact with the in-memory distributed data structure to process queries using the in-memory distributed data structure and to modify the in-memory distributed data structure. The data analysis system creates uniform resource identifier (URI) to identify each in-memory distributed data structure. The URI can be communicated from one client application to another application using communication mechanisms outside the data analysis system, for example, by email, thereby allowing other client devices to interact with a particular in-memory distributed data structure. The in-memory distributed data structure can be a machine learning model that is trained by one client device and executed by another client device. A client application can interact with the in-memory distributed data structure using different programming languages.

Abstract: Sequence data, such as time series data is analyzed using neural networks, for example, recurrent neural networks. The sequence data is obtained from a source. For example, a sequence data may represent time series data obtained from a sensor. As another example, the sequence of data may represent a sequence of user interactions performed by a user with an online system. The sequences of data are provided as input to a neural network. A feature vector representation of each input sequence data is extracted from the neural network. The feature vector representation is used for clustering the sequence data. Salient features of clusters of sequence data are determined. The salient features of clusters of sequence data are provided for display via a user interface.

Abstract: A data analysis system stores in-memory representation of a distributed data structure across a plurality of processors of a parallel or distributed system. Client applications interact with the in-memory distributed data structure to process queries using the in-memory distributed data structure and to modify the in-memory distributed data structure. The data analysis system creates uniform resource identifier (URI) to identify each in-memory distributed data structure. The URI can be communicated from one client application to another application using communication mechanisms outside the data analysis system, for example, by email, thereby allowing other client devices to interact with a particular in-memory distributed data structure. The in-memory distributed data structure can be a machine learning model that is trained by one client device and executed by another client device. A client application can interact with the in-memory distributed data structure using different programming languages.

Abstract: A data analysis system stores in-memory representation of a distributed data structure across a plurality of processors of a parallel or distributed system. Client applications interact with the in-memory distributed data structure to process queries using the in-memory distributed data structure and to modify the in-memory distributed data structure. The data analysis system creates uniform resource identifier (URI) to identify each in-memory distributed data structure. The URI can be communicated from one client application to another application using communication mechanisms outside the data analysis system, for example, by email, thereby allowing other client devices to interact with a particular in-memory distributed data structure. The in-memory distributed data structure can be a machine learning model that is trained by one client device and executed by another client device. A client application can interact with the in-memory distributed data structure using different programming languages.