| 摘要 |
In order to improve the prediction accuracy of a service level agreement breach, disclosed is an agreement breach prediction system provided with: a probability distribution calculation unit (21) which with reference to a learning event log (L1) and a learning SLA log (L2), calculates the probability distribution of a measured value for each event; a probability calculation unit (22) which with reference to the calculated probability distribution and an SLA list (L7), calculates the probability that falls into an SLA breach and the probability that does not fall into the SLA breach for each event; an SLA distribution calculation unit (23) which with reference to the learning SLA log (L2), calculates the probability density of the measured value that falls into the SLA breach during a predetermined period of time after the occurrence of the event and the probability density of the measured value that does not fall into the SLA breach during the predetermined period of time after the occurrence of the event; and a breach occurrence probability calculation unit (24) which when the event occurs, calculates the probability that the SLA breach occurs within the predetermined period of time after the occurrence of the event using the respective probabilities corresponding to the event and the respective probability densities of the measured value. |
