System and a method for training a neural network having autoencoder architecture to recover missing data

A system for training a neural network with an autoencoder architecture including an encoder for encoding its inputs into a latent space and a decoder for decoding the encodings from the latent space, comprising:

• at least one processor;
• and memory having instructions stored thereon that, when executed by the at least one processor, cause the system to:
• create a first training set including a valid data set of multiple dimensions, such that each valid instance of the valid data set includes multiple valid data points having values within a range of valid values of their corresponding dimensions;
• train the encoder and the decoder of the neural network in a first training stage using the first training set to reduce a difference between the valid data set provided to the encoder and a data set decoded by the decoder;
• create a second training set comprising an invalid data set formed by replacing at least some valid data points of the valid instances with invalid data points having values outside of the range of valid values of their corresponding dimensions;
• and
• train the encoder of the neural network in a second training stage using the second training set to reduce a difference between encodings of the valid data instances and encodings of their corresponding invalid data instances.

A data recovery system for producing a full set of measurements of a set of sensors from an incomplete set of measurements missing a measurement from at least one sensor in the set of sensors, comprising:

• at least one processor;
• and memory having instructions stored thereon that, when executed by the at least one processor, cause the data recovery system to:
• replace the missing measurement in the incomplete set of measurements with an invalid value outside of a range of valid values of the missing measurement to complete the incomplete set of measurements;
• process the completed set of measurements using an autoencoder to produce the full set of measurements with a recovered value of the missing measurement, wherein the autoencoder is a neural network learned based on a multi-stage training procedure;
• and
• output the full set of measurements with the recovered value of the missing measurement.