Application of kernel Fisher method based on primary factor analysis to recognition problem between oil layer and water layer
-
摘要: 根據測井數據結構復雜和交集嚴重的特點,將主成分分析思想應用到剔除奇異點和尋找兩類樣本的交集中,并在交集中應用核Fisher判別方法,進行油水判別,彌補了Fisher線性判別方法的不足.通過將主成分分析和核Fisher判別方法這兩種理論有機的結合起來,提高了利用測井數據識別油水層的鑒別能力,實際應用中證明了本方法的實用性和有效性.Abstract: The idea of primary component analysis was applied to eliminating the singular point and selecting the intersection of raw log data sets according to the characteristics of raw log data. Then kernel Fisher method was used in the intersection, which remedy the shortcoming of linear differentiate methods. By combining the two method, primary component analysis and kernel Fisher, the differentiate capability was improved and the practicability is testified in application.
-
Key words:
- primary factor analysis /
- singular point /
- kernel Fisher
-

計量
- 文章訪問數: 136
- HTML全文瀏覽量: 23
- PDF下載量: 6
- 被引次數: 0