Dynamic Data Fluctuation Analysis via Fitting Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring abnormal fluctuations in test data, such as website traffic, often fail to detect abrupt changes in a timely manner, leading to potential faults and losses, as they rely on static thresholds or historical comparisons without advanced statistical analysis.
Innovation Solution
A method and apparatus that acquire test data, perform fitting calculations to estimate data fluctuations, and generate data fluctuation values by comparing current and historical differences, using normal or exponential distributions to characterize and output these fluctuations, with optional warning triggers based on preset thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If static thresholds or simple historical comparison methods are used for monitoring test data, then the monitoring system is simple to implement, but the accuracy and timeliness of abnormal fluctuation detection deteriorates
Solution Approach 1:
The patent transforms the monitoring approach by changing from static threshold parameters to dynamic statistical parameters (mean, standard deviation) calculated from historical data. This allows the system to adapt to varying data patterns while maintaining detection accuracy, resolving the contradiction between simplicity and precision.
Solution Approach 2:
The patent introduces intermediate statistical calculations (fitting values, difference values, standard scores) as mediators between raw test data and abnormality detection. These intermediaries transform complex data analysis into a systematic process that improves detection accuracy without proportionally increasing system complexity.
2Measurement precision
If advanced statistical analysis methods are used to improve abnormal fluctuation detection accuracy, then detection precision improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the data analysis process into distinct computational stages: data collection, fitting calculation, difference calculation, and standard score calculation. This segmentation allows each stage to be optimized independently and enables efficient processing by breaking down complex computations into manageable steps.
Solution Approach 2:
The patent performs preliminary actions by pre-calculating statistical parameters (mean, standard deviation) from historical data and storing them for quick retrieval. This preliminary processing reduces the computational burden during real-time monitoring, maintaining high detection accuracy while reducing processing time.
3Measurement precision
If fitting calculations and statistical distributions are used to characterize data fluctuations, then the accuracy of fluctuation characterization improves, but the computational resources required increase
Solution Approach 1:
The system performs self-service by automatically calculating and updating statistical parameters from the test data itself without requiring external intervention or pre-configured complex models. The fitting calculations use the data's own historical patterns to generate prediction models, reducing the need for external computational resources.
Data Source
AI summary
A method and an apparatus for outputting information are provided. The method includes: acquiring at least one piece of test data generated in a target period; performing a fitting calculation on the at least one piece of test data to obtain estimated test data in the target period; and determining, for a test time point in the target period, a difference value between test data generated at the test time point and corresponding estimated test data as a target difference value; determining at least one historical correspondence difference value corresponding to the determined target difference value; and generating and outputting a data fluctuation value based on the determined target difference value and the determined at least one historical correspondence difference value.