Gradient Boosting Model for Real-Time Transaction Volume Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing transaction monitoring platforms struggle to detect real-time transaction volume anomalies in network services, such as mobile applications, which can lead to customer-facing issues and increased maintenance costs due to unnoticed errors and inefficiencies in tracking and responding to these anomalies.
Innovation Solution
A server device uses historical transaction data and historical calendar data to train a machine learning model with a gradient boosting technique to predict normal transaction volumes and confidence bands, enabling the detection of anomalies and generating alerts for real-time transaction volume deviations, thereby facilitating proactive responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional transaction monitoring platforms are used, then basic transaction tracking is possible, but real-time anomaly detection capability is insufficient
Solution Approach 1:
The system performs preliminary actions by training machine learning models in advance using historical transaction data and calendar data to establish baseline expectations for normal transaction volumes. This preliminary model training enables the system to detect anomalies in real-time without requiring complex real-time analysis, thus reducing the mean time to detect issues while improving reliability.
Solution Approach 2:
The patent replaces traditional mechanical monitoring approaches with machine learning-based detection. The machine learning model processes historical data patterns and automatically identifies anomalies in real-time transaction volumes, substituting manual or rule-based monitoring with intelligent automated detection that achieves both higher reliability and faster detection times.
2Productivity
If manual monitoring and response processes are used, then system complexity is lower, but productivity and response speed are reduced
Solution Approach 1:
The monitoring system performs self-service by automatically detecting anomalies and generating alerts without requiring manual intervention. The machine learning model continuously monitors transaction volumes, compares them against predicted baselines, and autonomously identifies deviations, thereby increasing productivity while the automated nature of the system manages its own complexity rather than requiring manual oversight.
Solution Approach 2:
The system implements feedback mechanisms where real-time transaction data is continuously fed back into the machine learning model for comparison against historical patterns. This feedback loop enables rapid anomaly detection and model refinement, improving response speed while the automated feedback processing manages system complexity through algorithmic rather than manual control.
3Measurement precision
If comprehensive real-time monitoring is implemented, then detection accuracy improves, but computing resource consumption increases
Solution Approach 1:
The system applies partial action by focusing monitoring resources on detecting significant anomalies rather than processing every transaction detail equally. The machine learning model identifies patterns and flaggs only the most relevant deviations from normal behavior, achieving high detection accuracy for critical anomalies while reducing overall computing resource consumption by avoiding exhaustive analysis of all transaction data.
Solution Approach 2:
The patent utilizes parameter changes by dynamically adjusting monitoring intensity based on predicted transaction volumes and anomaly probabilities. The machine learning model processes historical data to understand normal parameter variations and only performs detailed analysis when deviations are detected, maintaining high measurement precision for accuracy while optimizing computing resource consumption by avoiding continuous heavy processing during normal operation.
Data Source
AI summary
A server device obtains historical transaction data regarding transactions involving a network service, obtains historical calendar data regarding static date information for a historical time period that corresponds with the historical transaction data, and processes the historical transaction data and historical calendar data to train a machine learning model using a gradient boosting machine learning technique to predict a normal transaction volume for a period of time and confidence bands associated with the normal transaction volume. The server device generates the normal transaction volume for the period of time and confidence bands using the machine learning model, obtains real-time data concerning a transaction volume during the period of time, detects a transaction volume anomaly based on comparing the real-time data and normal transaction volume and confidence bands, and sends an alert, based on the transaction volume anomaly, to cause a remote device to display the alert and perform an action.