Processing event data and/or tabular data for input to one or more machine learning models

An apparatus configured to cluster and tokenize event data for machine learning classification tasks, the apparatus comprising:

• one or more processors;
• and
• memory storing executable instructions that, when executed by the one or more processors, cause the apparatus to:
• train, using training data, one or more machine learning models to perform a classification task, wherein the training data comprises a plurality of historical sequential data tokens corresponding to past events and corresponding indications of whether or not each of the historical sequential data tokens correspond to fraud;
• receive event data comprising, for each event of a plurality of different events, a plurality of corresponding data fields;
• identify a plurality of different clusters for the event data by separating, based on a target quantity of clusters and based on values of the plurality of corresponding data fields, each of a plurality of different portions of the event data into a corresponding cluster of the plurality of different clusters;
• generate a grid-based version of the plurality of different clusters by:
• defining a grid comprising a plurality of cells, wherein a quantity of the plurality of cells is based on a predetermined level of granularity for the event data:
• assigning each cluster of the plurality of different clusters to a different cell of the plurality of cells;
• and
• assigning a plurality of token values by assigning, to each of the plurality of cells that is associated with at least one cluster of the plurality of different clusters, a different token value, of the plurality of token values, representing a category of the event data;
• receive first event data corresponding to a first event;
• generate a sequential data token by:
• simplifying, based on the predetermined level of granularity, a plurality of different data values of the first event data;
• determining, by comparing the simplified plurality of different data values of the first event data to the grid-based version of the plurality of different clusters, a first cell, of the plurality of cells, corresponding to the first event data;
• and
• generating the sequential data token based on one or more tokens assigned to the first cell;
• provide, as input to a first language model, the sequential data token;
• receive, as output from the first language model, first generative text generated based on the sequential data token;
• provide, as input to the one or more machine learning models trained to perform the classification task, the first generative text;
• and
• receive, as output from the one or more machine learning models, a classification of the first event.

One or more non-transitory computer-readable media storing executable instructions that, when executed, cause one or more computing devices to cluster and tokenize event data for machine learning classification tasks by causing the one or more computing devices to:

• train, using training data, one or more machine learning models to perform a classification task, wherein the training data comprises a plurality of historical sequential data tokens corresponding to past events and corresponding indications of whether or not each of the historical sequential data tokens correspond to fraud;
• receive event data comprising, for each event of a plurality of different events, a plurality of corresponding data fields;
• identify a plurality of different clusters of the event data by separating, based on a target quantity of clusters and based on values of the plurality of corresponding data fields, each of a plurality of different portions of the event data into a corresponding cluster of the plurality of different clusters;
• generate a grid-based version of the plurality of different clusters by:
• defining a grid comprising a plurality of cells, wherein a quantity of the plurality of cells is based on a predetermined level of granularity for the event data;
• assigning each cluster of the plurality of different clusters to a different cell of the plurality of cells;
• and
• assigning a plurality of token values by assigning, to each of the plurality of cells that is associated with at least one cluster of the plurality of different clusters, a different token value, of the plurality of token values, representing a category of the event data;
• receive first event data corresponding to a first event;
• generate a sequential data token by:
• simplifying, based on the predetermined level of granularity, a plurality of different data values of the first event data;
• determining, by comparing the simplified plurality of different data values of the first event data to the grid-based version of the plurality of different clusters, a first cell, of the plurality of cells, corresponding to the first event data;
• and
• generating the sequential data token based on one or more tokens assigned to the first cell;
• provide, as input to a first language model, the sequential data token;
• receive, as output from the first language model, first generative text generated based on the sequential data token;
• provide, as input to the one or more machine learning models trained to perform the classification task;
• and
• receive, as output from the one or more machine learning models, a classification of the first event.