Abstract: A device includes one or more processors configured to process first input time-series data associated with a first time range using an embedding generator to generate an input embedding. The input embedding includes a positional embedding and a temporal embedding. The positional embedding indicates a position of an input value within the first input time-series data. The temporal embedding indicates that a first time associated with the input value is included in a particular day, a particular week, a particular month, a particular year, a particular holiday, or a combination thereof. The processors are configured to process the input embedding using a predictor to generate second predicted time-series data associated with a second time range. The second time range is subsequent to at least a portion of the first time range. The processors are configured to provide, to a second device, an output based on the second predicted time-series data.

Abstract: A method includes obtaining sensor data associated with operation of one or more devices and providing input data based on the sensor data to a dimensional-reduction model that includes a first layer having a first count of nodes, a second layer having a second count of nodes, and a third layer having a third count of nodes. The second layer is disposed between the first layer and the third layer, and the second count of nodes is greater than the first count of nodes and the third count of nodes. The method also includes determining a reconstruction error indicating a difference between the input data and the output data of the dimensional-reduction model. The method also includes performing a comparison of the reconstruction error to an anomaly detection criterion and generating an anomaly detection output for the one or more devices based on a result of the comparison.

Abstract: A method includes obtaining sensor data associated with operation of one or more devices and providing input data based on the sensor data to a dimensional-reduction model that includes a first layer having a first count of nodes, a second layer having a second count of nodes, and a third layer having a third count of nodes. The second layer is disposed between the first layer and the third layer, and the second count of nodes is greater than the first count of nodes and the third count of nodes. The method also includes determining a reconstruction error indicating a difference between the input data and the output data of the dimensional-reduction model. The method also includes performing a comparison of the reconstruction error to an anomaly detection criterion and generating an anomaly detection output for the one or more devices based on a result of the comparison.

Abstract: A method includes obtaining sensor data associated with operation of one or more devices and providing input data based on the sensor data to a dimensional-reduction model having an encoder portion and a decoder portion and configured such that the encoder portion is not mirrored by the decoder portion. The method also includes obtaining output data from the dimensional-reduction model responsive to the input data and determining a reconstruction error indicating a difference between the input data and the output data. The method also includes performing a comparison of the reconstruction error to an anomaly detection criterion and generating an anomaly detection output for the one or more devices based on a result of the comparison.

Abstract: A method includes, during an epoch of a genetic algorithm, determining a fitness value for each of a plurality of autoencoders. The fitness value for an autoencoder indicates reconstruction error responsive to data representing a first operational state of one or more devices. The method includes selecting, based on the fitness values, a subset of autoencoders. The method also includes performing a genetic operation with respect to at least one autoencoder to generate a trainable autoencoder. The method includes training the trainable autoencoder to reduce a loss function value to generate a trained autoencoder. The loss function value is based on reconstruction error of the trainable autoencoder responsive to data representative of a second operational state of the device(s). The method includes adding the trained autoencoder to a population to be provided as input to a subsequent epoch of the genetic algorithm.

Abstract: A method includes, during an epoch of a genetic algorithm, determining a fitness value for each of a plurality of autoencoders. The fitness value for an autoencoder indicates reconstruction error responsive to data representing a first operational state of one or more devices. The method includes selecting, based on the fitness values, a subset of autoencoders. The method also includes performing a genetic operation with respect to at least one autoencoder to generate a trainable autoencoder. The method includes training the trainable autoencoder to reduce a loss function value to generate a trained autoencoder. The loss function value is based on reconstruction error of the trainable autoencoder responsive to data representative of a second operational state of the device(s). The method includes adding the trained autoencoder to a population to be provided as input to a subsequent epoch of the genetic algorithm.