Identification of anomalous transaction attributes in real-time with adaptive threshold tuning

A method, by a system comprising a hardware processor, configured to perform an ordered combination of actions, comprising:

• evaluating, by the system, a set of consumer transactions conducted during a defined time period, wherein the set includes one or more attributes that include at least a customer infrastructure characteristic and an insufficient transaction activity evaluation;
• categorizing, by the system, transactions within the set of consumer transactions into defined groups;
• removing, by the system, outlier transactions from the set of consumer transactions;
• identifying, by the system, a set of non-anomalous transactions;
• receiving, by the system, a subsequent consumer transaction;
• and
• performing, by an adaptive tuning processor of the system, the following:
• automatically allowing the subsequent transaction based on a determination that the subsequent transaction conforms to the set of non-anomalous transactions;
• or
• automatically rejecting the subsequent transaction and outputting an alert for further analysis of the subsequent transaction based on another determination that the subsequent transaction does not conform to the set of non-anomalous transactions,
• wherein the automatically allowing or automatically rejecting integrates a consumer ceiling threshold via a modeling based on at least a modeled transaction attribute and the subsequent transaction by individually tuning a plurality of segments of the modeled transaction attributes on an ongoing basis;
• and
• wherein the ordered combination of actions mitigate computing fraud.

A computer-readable storage device that stores executable instructions that, in response to execution, cause a system comprising a hardware processor to configure the hardware processor to perform operations, comprising:

• systematically modeling one or more attributes of transactions of a consumer included in a set of historical transactions of the consumer, wherein the one or more attributes include at least a customer infrastructure characteristic and an insufficient transaction activity evaluation;
• determining a consumer ceiling threshold for one or more transaction level characteristics associated with the modeling, the one or more transaction level characteristics represent respective values beyond which a subsequent transaction of the consumer is considered anomalous for that characteristic;
• providing real-time, or substantially real-time, detection of anomalous transactions and non-anomalous transactions, wherein respective alerts are output based on detection of the anomalous transactions;
• and
• providing systematic processing one of an allowance or a rejection of the anomalous and non-anomalous transactions based at least in part on the respective alerts,
• wherein the systematic processing integrates the consumer ceiling threshold via the modeling to base the one of allowance or rejection on at least the modeled transaction attributes and the subsequent transaction by individually tuning a plurality of segments of the modeled transaction attributes on an ongoing basis;
• and
• wherein the components execute to mitigate computing fraud in a computing environment.