Abstract: A neural network embodiment comprises an input layer, an output layer and a filter layer. Each unit of the filter layer receives a filter layer input from a single preceding unit via a respective filter layer input connection. Each filter layer input connection is coupled to a different single preceding unit. The filter layer incentivizes the neural network to learn to produce a target output from the output layer for a given input to the input layer while simultaneously learning weights for each filter layer input connection. The weights learned cause the filter layer to reduce a number of filter layer units that pass respective filter layer inputs as non-zero values. When applied as an initial internal layer between an input layer and an output layer, the filter layer incentivizes the neural network to learn which neural network input features to discard to produce the target output.

Abstract: A neural network embodiment comprises an input layer, an output layer and a filter layer. Each unit of the filter layer is configured to receive a filter layer input from a single preceding unit via a respective filter layer input connection. Each filter layer input connection is coupled to a different single preceding unit. The filter layer is configured to incentivize the neural network to learn to produce a target output from the output layer for a given input to the input layer while simultaneously learning weights for each filter layer input connection. The weights learned cause the filter layer to reduce a number of filter layer units that pass respective filter layer inputs as non-zero values. When applied as an initial internal layer between an input layer and an output layer, the filter layer incentivizes the neural network to learn which neural network input features to discard to produce the target output.

Abstract: According to some aspects, a system for automatically processing text comprising information regarding a patient encounter to assign medical codes to the text is provided. The system comprises at least one storage medium storing processor-executable instructions, and at least one processor configured to execute the processor-executable instructions to perform analyzing the text to extract a plurality of facts from the text, identifying at least one of the plurality of facts to be excluded from consideration when assigning medical codes to the text, and evaluating each of the plurality of facts, except for the identified at least one fact, to assign one or more medical codes to the text.

Abstract: A method and system of removing noise from a dictionary using a weighted graph is presented. The method can include mapping, by a noise reducing agent executing on a processor, a plurality of dictionaries to a plurality of vertices of a graphical representation, wherein the plurality of vertices is connected by weighted edges representing noise. The plurality of dictionaries may further comprise a plurality of entries, wherein each entry further comprises a plurality of tokens. The method can include selecting a subset of the weighted edges, constructing an acyclic graphical representation from the selected subset of weighted edges, and determining an ordering based on the acyclic graphical representation. The selected subset of weighted edges may approximate a solution to the Maximum Acyclic Subgraph problem. The method can include removing noise from the plurality of dictionaries according to the determined ordering.

Abstract: A method and system of removing noise from a dictionary using a weighted graph is presented. The method can include mapping, by a noise reducing agent executing on a processor, a plurality of dictionaries to a plurality of vertices of a graphical representation, wherein the plurality of vertices is connected by weighted edges representing noise. The plurality of dictionaries may further comprise a plurality of entries, wherein each entry further comprises a plurality of tokens. The method can include selecting a subset of the weighted edges, constructing an acyclic graphical representation flom the selected subset of weighted edges, and determining an ordering based on the acyclic graphical representation. The selected subset of weighted edges may approximate a solution to the Maximum Acyclic Subgraph problem. The method can include removing noise from the plurality of dictionaries according to the determined ordering.