Integrated machine learning and rules platform for improved accuracy and root cause analysis

A method for machine learning and rules integration, comprising:

• receiving, from a user via a user interface, a plurality of input values corresponding to a subset of a set of input variables associated with an automated determination, wherein the subset does not include at least one input variable from the set of input variables;
• generating a directed acyclic graph (DAG) representing a set of constraints corresponding to the set of input variables, wherein the set of constraints relate to:
• one or more machine learning models;
• and
• one or more rules;
• providing one or more inputs to a subset of the one or more machine learning models based on one or more input values of the plurality of input values;
• receiving one or more outputs from the subset of the one or more machine learning models based on the one or more inputs;
• determining outcomes for the one or more rules based on at least one input value of the plurality of input values;
• populating the DAG based on the plurality of input values, the one or more outputs from the subset of the one or more machine learning models, and the outcomes for the one or more rules;
• determining not to utilize a given machine learning model of the one or more machine learning models based on logic represented by the DAG;
• making the automated determination based on the logic represented by the DAG;
• providing information for display to the user via the user interface based on the automated determination;
• receiving, from the user via the user interface, feedback relating to the information;
• using the DAG to identify a particular machine learning model of the one or more machine learning models to be re-trained based on the feedback, wherein the particular machine learning model is re-trained through a supervised learning process comprising:
• providing one or more training inputs to the particular machine learning model based on the plurality of input values;
• receiving one ore more training outputs from the particular machine learning model based on the one or more training inputs;
• and
• iteratively modifying one ore more parameters of the particular machine learning model based on the one ore more training outputs and the feedback;
• and
• making a subsequent automated determination based on the re-trained particular machine learning model.

A system, comprising:

• one or more processors;
• and
• a memory comprising instructions that, when executed by the one or more processors, cause the system to:
• receive a plurality of input values corresponding to a subset of a set of input variables associated with an automated determination, wherein the subset does not include at least one input variable from the set of input variables;
• generate a directed acyclic graph (DAG) representing a set of constraints corresponding to the set of input variables, wherein the set of constraints relate to:
• one or more machine learning models;
• and
• one or more rules;
• provide one or more inputs to a subset of the one or more machine learning models based on one or more input values of the plurality of input values;
• receive one or more outputs from the subset of the one or more machine learning models based on the one or more inputs;
• determine outcomes for a subset of the one or more rules based on at least one input value of the plurality of input values;
• populate the DAG based on the plurality of input values, the one or more outputs from the subset of the one or more machine learning models, and the outcomes for the subset of the one or more rules;
• determine not to utilize a given machine learning model of the one or more machine learning models based on logic represented by the DAG;
• make the automated determination based on the logic represented by the DAG;
• perform one or more actions based on the automated determination;
• receive user feedback with respect to the automated determination;
• and
• use the DAG to identify a particular machine learning model of the one or more machine learning models to be re-trained based on the user feedback, wherein the particular machine learning model is re-trained through a supervised learning process comprising:
• providing one or more training inputs to the particular machine learning model based on the plurality of input values;
• receiving one ore more training outputs from the particular machine learning model based on the one or more training inputs;
• and
• iteratively modifying one ore more parameters of the particular machine learning model based on the one or more training outputs and the user feedback.