Model management system for developing machine learning models

A computer program product for developing a geographic agnostic machine learning model, the computer program product comprising at least one non-transitory computer-readable medium including one or more instructions that, when executed by at least one processor, cause the at least one processor to:

• select transaction data associated with payment transactions conducted by a first plurality of users, wherein the transaction data comprises first transaction data associated with payment transactions conducted by a first plurality of users in a first geographic area and second transaction data associated with payment transactions conducted by a second plurality of users in a second geographic area;
• format the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area to provide geographic agnostic transaction data associated with payment transactions;
• wherein, the one or more instructions that cause the at least one processor to format the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area, cause the at least one processor to:
• normalize the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area;
• wherein, the one or more instructions that cause the at least one processor to normalize the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area, cause the at least one processor to:
• perform a de-weighting technique on the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area;
• wherein, the one or more instructions that cause the at least one processor to perform the de-weighting technique, cause the at least one processor to:
• perform the de-weighting technique based on:
• a parameter associated with whether an account is classified as an account that is associated with a user that travels between one geographic area and another geographic area, or
• a parameter associated with whether an account is classified as an account that is associated with a user that is likely to close the account within a predetermined period of time;
• generate a geographic agnostic machine learning model using training data, wherein the training data comprises the geographic agnostic transaction data associated with payment transactions, and wherein the geographic agnostic machine learning model is configured to provide an output that includes a prediction that an account associated with an input should be assigned to a category of a plurality of categories;
• generate a generalization layer based on data associated with macro-economic factors of the first geographic area and data associated with macro-economic factors of the second geographic area;
• apply an input to the generalization layer to provide an output of the generalization layer;
• and
• provide the output of the generalization layer as an input to the geographic agnostic machine learning model.

A computer-implemented method for developing a geographic agnostic machine learning model, the method comprising:

• selecting, with at least one processor, transaction data associated with payment transactions conducted by a first plurality of users, wherein the transaction data comprises first transaction data associated with payment transactions conducted by a first plurality of users in a first geographic area and second transaction data associated with payment transactions conducted by a second plurality of users in a second geographic area;
• formatting, with at least one processor, the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area to provide training data;
• wherein formatting the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area comprises:
• normalizing the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area;
• wherein normalizing the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area comprises:
• performing a de-weighting technique on the first transaction data associated with payment transactions conducted by the first plurality of users in the first geographic area and the second transaction data associated with payment transactions conducted by the second plurality of users in the second geographic area;
• wherein performing the de-weighting technique comprises:
• performing the de-weighting technique based on:
• a parameter associated with whether an account is classified as an account that is associated with a user that travels between one geographic area and another geographic area, or
• a parameter associated with whether an account is classified as an account that is associated with a user that is likely to close the account within a predetermined period of time;
• generating, with at least one processor, a geographic agnostic machine learning model using the training data, wherein the geographic agnostic machine learning model is configured to provide an output that includes a prediction that an account associated with an input should be assigned to a category of a plurality of categories;
• and
• generating a generalization layer based on data associated with macro-economic factors of the first geographic area and data associated with macro-economic factors of the second geographic area.