Systems and methods for improved machine learning models

A method comprising:

• receiving, by an application executed by one or more processors, indications comprising a beneficiary account identifier, a beneficiary bank identifier, and/or a beneficiary bank country;
• conducting, by the application, pre-processing on the beneficiary account identifier, the beneficiary bank identifier, and/or the beneficiary bank country;
• selecting, by the application, multiple features associated with the beneficiary account identifier, the beneficiary bank identifier, and/or the beneficiary bank country;
• conducting, by the application, a first Bayesian hyperparameter optimization on a grid space of hyperparameters to select hyperparameter values of the hyperparameters, wherein the hyperparameters includes an iteration amount;
• training, by the application, a machine learning model executed by one or more processors using the multiple features and the hyperparameter values, the machine learning model being trained on an indication training set;
• removing, by the application, one or more features of the multiple features to form a final feature set, the one or more features to be removed being determined by the application based on a SHAPley analysis of importance of each feature;
• conducting, by the application, a second Bayesian hyperparameter optimization on the grid space to select updated hyperparameter values associated with the hyperparameters;
• validating, by the application, the machine learning model using the updated hyperparameter values and the final feature set on a validation set;
• determining, by the machine learning model, a fraud based on the indications;
• and
• blocking, by a transaction application in communication with the application, the transaction based on the machine learning model determination.