This note is for chapters from 15 to 19, around the topics of 3 advanced algorithm analysis/design strategies and two relatively complex data structures “B-tree” & “Binomial-Heap”.
Part 4 containing chapters from 15 through 17 elaborates 3 very useful analysis/design techniques: dynamic programming, greedy strategy (one extension of dynamic programming), amortized analysis. I have got the gist of this part, but extra study on these topics are still necessary of course. You have to apply them in some application contexts to make them clearer.
Another two chapters are in the section of “advanced data structures” including B-Tree and Binomial Heap. In my opinion these two structures are more complex trees, but the basic ideas are not strange. Actually these two chapters remind me the classical R-B Tree, because the basic idea of the two structures still emphasizes how to maintain the basic natures of the structures. One interesting point is that the naming of B-Tree is still not clear... as http://en.wikipedia.org/wiki/B-Tree#Etymology_unknown
There are still two interesting chapters in Part 5: Fibonacci Heap and Data Structures for Disjoint Sets. And the last rudiment part is Graph Algorithms. I guess it will be even more fun.