Sub-User Data Change Detection for Server Communication Status
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to effectively detect changes in sub-user data for determining the condition of network communication between servers, especially when the output distribution is unknown, making it difficult to identify anomalies and maintain reliable data communication.
Innovation Solution
A computer-implemented system that includes a user server and a platform server, which detects sub-user data, calculates thresholds, rolling means, and cumulative sums to identify anomalies in communication states, and feeds back detected conditions to the user server for corrective action.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If change detection is performed on sub-user data with unknown output distribution, then anomaly detection capability is improved, but measurement precision deteriorates due to difficulty in determining normal state
Solution Approach 1:
The system performs preliminary actions by calculating rolling means and establishing thresholds before actual anomaly detection. The platform server calculates rolling means for each sub-period and establishes thresholds based on historical data patterns, enabling accurate anomaly detection even when the output distribution is initially unknown. This preliminary setup creates a baseline for comparison that improves subsequent detection precision.
Solution Approach 2:
The system employs parameter changes by dynamically adjusting thresholds based on rolling mean calculations. The thresholds are not fixed but adapt to changing data patterns through continuous recalculation of rolling means and standard deviations. This allows the system to maintain high detection accuracy while adapting to unknown or evolving output distributions.
2Measurement precision
If rolling mean calculation is performed for each sub-period to improve detection accuracy, then measurement precision is improved, but computational complexity increases
Solution Approach 1:
The system applies segmentation by dividing the data stream into sub-periods and calculating rolling means for each segment separately. This segmentation allows the platform server to process data in manageable chunks, improving detection accuracy by capturing local patterns while reducing the computational burden compared to processing the entire data stream at once. Each sub-period can be independently analyzed and then aggregated for final anomaly detection.
3Difficulty of detecting and measuring
If cumulative sum accumulation is performed to detect anomalies, then anomaly detection capability is improved, but loss of time increases due to data processing requirements
Solution Approach 1:
The system implements continuity of useful action by performing cumulative sum accumulation in a continuous, incremental manner rather than batch processing. The platform server maintains running cumulative sums as data arrives, allowing anomaly detection to occur with minimal delay. This continuous update approach reduces processing time by avoiding repeated full-data recalculations and enables real-time or near-real-time anomaly detection.
Data Source
AI summary
Apparatuses, methods, and systems for detecting changes in periodic sub-user behavior data for determining a condition of a network are disclosed. One method includes sensing the user data of sub-users, calculating a plurality of thresholds based on a size of the user data, calculating a rolling mean for each sub-period of a larger period, generating difference values based on comparing current sub-period values to corresponding sub-period rolling mean, accumulating a cumulative sum based the difference values, detecting an anomaly based on comparing the cumulative sum with the plurality of thresholds, detecting a condition of a state of communication between the user server and the platform server when the comparing of the cumulative sum with the plurality of threshold exceeds a condition detection threshold, and feeding the condition of the state of communication back to the user server when the condition is detected.