Diablo Forum

When I looked at it, I didn't see it; When it was removed, I felt its existence; When it was brought in, I realized its significance; Finally I carve its figure in my heart and then I get a visual; This is called recognition.

From time to time, I was bothered by a simple question. Which way is easier for us to notify? To add a small tree into a forest or to remove a small tree from a forest? If this is considered as a meaningless question, then how about this one? To add a tree in a grassland where no tree is there at all or to remove the single tree from grassland so there is no tree in grassland. Which way is easier to notice? Probably same and most people would say that.

If the above questions are all nonsense, then how about this question. In a structured system, say an AVL tree which means a balanced binary search tree. Does it cost same when you either add one element into the tree or remove one element from the tree? Not exactly same because it depends how we implement the structure of AVL tree. In some data structure, maybe removing is easier. So, back the above question, it depends. It depends on how you construct your knowledge system to represent what you see and how you implement a "fast-recognition" algorithm.

Of course the above comments are all garbage which I don't believe myself. I finished the "regular-sampling-quick-sorting" of MPI.

14/11/06

Yesterday Dr. Hopcroft came to make a lecture and here is some inspiration from his lecture. Or precisely it is simply what I remembered.

24/12/06

I guess the origin of AI is from vision because it is the most intuitive method to observe and recognize pattern. Maybe all high-level concepts need to be first transferred into visual image before comparison and recognition are done.

4/01/07

Convex Hull is a typical algorithm that the definition gives out the clue of construction itself! See, the definition says that any edge of the two vertices in the convex polygon will leave all other vertices in its right-hand-side. And the algorithm takes advantage of this. After the lecture I spent about one and half hour to finish a simple implementation of this algorithm. When I finished, I noticed the name of algorithm is "Graham's Scan-type". I am happy because it proves I am still capable of coding after so many months. Even though I am doing the MPI programming which is far more hard than these kind of single-machine programming.

14/01/07

99% of Chinese ancient philosophy is purely mysticism which is practically useless except for fortune-telling. Their dirty trick is to use as many as ambiguous terms, words, concepts, expressions etc. to express something so ambiguous such that they can always explain themselves if you are foolish enough to give the truth before asking for prediction. Most defenders of so-called Chinese traditions feel so blindly proud for this kind of useless philosophy and claim it to be unique. Let me quote President Bush to comment on this: Shame on you and a fool cannot be fooled again. (These are so popular in "youtube" and surely you will find it by searching "Bush") However, there is something I also feel like to note down and it is one of my favorite. For those who consider themselves as guru of programming and experts of AI, this might be interesting. "Expressions themselves cannot be expressed; Names themselves cannot be named." This is the very first sentence from <DaoDeJing> which is one of most famous ancient Chinese philosophy. And maybe it touches the most profound and most basic topics of language and origin of intelligence.

01/02/07

It is said that I think, so I live. In order to prove I am alive, I write a simple algorithm to generate simple polygon from a set of points because I only think when I do coding. Thus I live when I am programming. The trivial algorithm is inspired by assignment from comp6711.