| 摘要 |
The present invention relates to a method for detecting abnormality observation data using a time series prediction model and a method for detecting abnormality observation data regarding a groundwater level and, more specifically, to determining abnormality observation data based on results predicted by using past observation data. In particular, the present invention can determine abnormality observation data in real time by comparing predictions based on observation data measured in the past with predictions based on past prediction data using a support vector machine (SVM) algorithm. More specifically, a groundwater level time series prediction model can be created by using weather data such as precipitation, etc. having influence on a natural change of a groundwater level and groundwater level observation data, and the ability to detect abnormality observation data regarding a groundwater level can be improved by using the created groundwater level time series prediction model. Therefore, reliability and competitiveness can be improved in a natural resource management field associated with various weather changes and fields associated with or similar to the natural resource management field as well as a water resource management field, especially, a groundwater management field. |
