How to build a heap better than Incremental and In-place method using decision tree?

For an array A = [a1, a2, a3, a4] of distinct numbers, I have built heap using binary decision tree by Incremental and In-place method.

Incremental method:

In-place method:

Is there a way to build heap using decision tree for inputs of size 4 that uses fewer decisions in the worst-case, compared to above methods?

Notes on heap: https://drive.google.com/open?id=0By6GDPYLwp2cY3lfbEVWNHlrSlE Incremental method - page 36, In-place method - page 48.

• I don't think I understand what you're asking. Could you be more specific? – quicksort Feb 11 '17 at 15:29
• Every branch in above methods makes a comparison to built a heap. So, is there any other algo. that makes fewer comparison to built a heap? – New_Coder Feb 11 '17 at 15:42
• When $a_1$ and $a_3$ are less-than $a_2$ and $a_4$, your upper decision tree won't compare $a_1$ to $a_3$, so are your leaves just a rough idea of the heap? ​ ​ – user12859 Feb 11 '17 at 16:05