System and method for removing biases within a distributable model

A system for improving a distributable model with distributed data, comprising:

• a computing system comprising a processor, a memory, and a network interface, a directed computational graph subsystem comprising a plurality of programming instructions stored in the memory and operable on the processor, wherein the programming instructions, when operating on the processor, cause the computing system to:
• receive an instance of a distributable model from a distributable model source via a computer network;
• operate a directed computational graph comprising a plurality of data transformation nodes configured to process model data stored on the computing system by removing data biases wherein each of the plurality of data transformation nodes executes code to perform one or more data transformations on input data and outputs transformed data;
• train the instance of the distributed model on the model data to remove a plurality of data biases within the instance of the distributable model using the directed computational graph;
• create a cleansed dataset based on the trained instance of the distributable model and the removed biases;
• wherein the biases comprise trends exhibited within the data;
• wherein the biases are automatically intelligently weighted and corrected by the directed computational graph to generalize the dataset for use in a distributable model;
• train the instance of the distributable model with the cleansed dataset;
• and
• generate an update report based upon which remote instances of the distributable model may be adjusted to take into account the training of the instance of the distributable model using the cleansed dataset.