Skip to main content
Figure 5 | BMC Medical Genomics

Figure 5

From: Using logistic regression to improve the prognostic value of microarray gene expression data sets: application to early-stage squamous cell carcinoma of the lung and triple negative breast carcinoma

Figure 5

The revolving sliding window method for selection of 40 consecutive sets of 20 cases each. Shown here is the revolving sliding window for early deaths (the same approach was used with the 40 longest survivals). The first window selected is from case 1 to case 20, then from case 2 to 21, etc., until there are no longer 20 sequential cases available without going beyond the 40th case shown in Figure 2. At that point, the first earliest death case is used to complete the window of 20 and so on, using sequential cases. The 40 revolving sliding windows of early deaths or long survivors was compared to the opposing fixed group of 20 cases, also indicated in Figure 2. In total, 80 comparisons were made.

Back to article page