Correlation associating method for IT operation and maintenance indexes
A correlation and index technology, applied in the field of IT operation and maintenance management, can solve the problems of field strength signal-to-noise ratio strength affecting network data bandwidth, WEB authentication access delay time increase, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
specific Embodiment 2
[0066] In specific embodiment 2, please refer to the description figure 2 .
[0067] Assuming that in a real scenario, there are 10 indicators and the total number of training data is 10,000. Divide them into 100 groups with 100 data in each group. The training data source is a random number between 0 and 1, and the preset index values to be tested are from 0 increments by 0.5 until it approaches 20, set 1 to alarm (exceeding standard), 0 does not alarm. Theoretically, when the data is farther away from this range of 0 to 1, the alarm should be 1, otherwise it should be 0. Due to the robustness of the algorithm given by the method, after smoothing, the forecasting performance is reflected by the alarm probability.
[0068] By description figure 2 As shown, when the preset data (test data to be predicted) gradually moves away from 1, the alarm probability gradually rises until it approaches 1. In practice, the solution is to set up a threshold. When the alarm probability obtai...
specific Embodiment 3
[0070] In specific embodiment 3, please refer to the description image 3 .
[0071] Assuming that in a real scenario, the number of indicators is 20, and each set of indicator training data sources has 1000 data. The indicator data source to be tested is a random number with 10 as the mean and 0.1 as the variance. The preset test data for the indicators to be tested is from 5 to 5 Increment from 0.5 to 15, calculate the forecast deviation rate.
[0072] By description image 3 As shown, when the preset range is in the range of 10, the minimum prediction error can be lower than 0.1, otherwise, the prediction error will become larger and larger. This explanatory diagram, like the first embodiment, illustrates that the prediction method provided by the present invention has higher accuracy.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 