MCC
A classification evaluation metric that jointly accounts for true positives, true negatives, false positives and false negatives. Its value ranges from -1 to 1, and it is more robust than accuracy under class imbalance.
Timeline (last 90 days)
Under the recording-disjoint evaluation protocol, DySR-GNN achieves a recording-level Macro-F1 of 0.9524, an MCC of 0.9302, and a strict exact-match accuracy of 0.8519.
Merchant Category Code (MCC) is the technical basis of credit card rewards rules, with card networks assigning each merchant a four-digit code representing its industry category.
Issuers' bonus points rules are usually tied directly to merchant category codes rather than merchant brands.
Because of non-standard entry by acquirers or mixed merchant businesses, MCC matching is error-prone, which is one of the main technical challenges facing automated recommendation tools.
The Matthews Correlation Coefficient (MCC) is more reliable than accuracy and F1 score in class-imbalanced scenarios because it simultaneously considers true positives, false positives, true negatives, and false negatives.
The Matthews Correlation Coefficient (MCC) ranges from -1 to +1, where +1 indicates perfect prediction, 0 indicates random guessing, and -1 indicates completely inverse prediction
All Facts (6)
Under the recording-disjoint evaluation protocol, DySR-GNN achieves a recording-level Macro-F1 of 0.9524, an MCC of 0.9302, and a strict exact-match accuracy of 0.8519.
50%UnverifiedMerchant Category Code (MCC) is the technical basis of credit card rewards rules, with card networks assigning each merchant a four-digit code representing its industry category.
50%UnverifiedIssuers' bonus points rules are usually tied directly to merchant category codes rather than merchant brands.
50%UnverifiedBecause of non-standard entry by acquirers or mixed merchant businesses, MCC matching is error-prone, which is one of the main technical challenges facing automated recommendation tools.
50%UnverifiedThe Matthews Correlation Coefficient (MCC) is more reliable than accuracy and F1 score in class-imbalanced scenarios because it simultaneously considers true positives, false positives, true negatives, and false negatives.
50%UnverifiedThe Matthews Correlation Coefficient (MCC) ranges from -1 to +1, where +1 indicates perfect prediction, 0 indicates random guessing, and -1 indicates completely inverse prediction
50%