11/01/09: The 80-20 Rule

Reading Joshua Bloch's Effective Java (2nd Ed.) book today on item 55: Optimize judiciously He states: "...the part of the program that you think is slow may not be at fault, in which you would be wasting time trying to optimize it. Common wisdom says that programs spend 80 percent of their time in 20 percent of their code." I was amused to see that the 80-20 rule seems to be everywhere. I have heard of it in the past when a co-worker complained that he was one of the 20% of the employees who were responsible for 80% of the work while his paycheck didn't reflect the ratio. Then I learned about this rule in my management training course, which states that 80% of the problems are originated from 20% of the causes, so identifying and fixing those issues will be most cost effective. I think this is so true. I also observed that 80% of my teaching effort is often spent on 20% of my students. Today I googled and found more info on this rule.

10/26/09: Market Efficiency & Algorithm Trading

I bought some small new ear rings Sunday and put them on this morning. After a long day from work I was glad to see my two devils at home. "Mommy, you have new ear rings!", my six year old spotted right away. What an observing kid, I said to myself proundly. Then ran in my two year old and he gave me a big hug, saying: "Mommy, you look so nice in these new earings!". Wow, I was more surprised than flattered. Then my husband came home and we had dinner and had reading time and then story time. He never noticed my new ear rings. This made me think can the market be truly efficient?

I think NO. Although we live in an information age, the ability that people discover, process, and consume information can be totally differnt. My six year old discovered the information by stating the fact; My two year old discovered and consumed it because he gave his opinion on that fact; My husband didn't discover the information at all, probably he was tired, or he was faced with too much information around him. So even if we did have a pure, complete efficient market, information users' behavior will always be different and the consequent results will be different. That's why there are so many different fund companies with their proprietary trading algorithms.

But can computer algorithm discover and consume market information more effciently than a well trained human trader? I hope the answer is YES, but honestly I don't know. On one side, computer can process a vast amount information quickly and accurately. That's why Deep Blue can Kasparov. On the other side, there are factors/rules/chemstry or whatever you call them, that machine does not have, and more importantly machine cannot learn to have. Sometimes even humans cannot learn to have. For example, my husband is not and will never be a people person. The difference between a businessman and an entrepreneuars is more than just their IQ, but also their EQ. Certain people have certain unquantifiable qualities that make them successful. Sensibility is sometimes more important than sense.


10/18/09: Learning a New Language

We had a parent and teacher conference today for Robert's Chinese language school. Most parents agreed that the kids do not want to learn Chinese. Some even hate Chinese. So why should we force our kids to learn Chinese?

Well, let's look back and see how we were forced to learn English as a second language when we were kids. I liked learning English because I wanted to know the western cultures, to read English classic novels, to understand Michael Jackon's songs, to travel abroad. Language is a tool that helps me explore the world. So my next big project is to help my kids to see the power of knowing an additional language.

There are also some issues with language education in the old traditional methods. It's always boring - memorizing the spelling, reading the text book drills over and over, and the content of the textbook doesn't apply well to daily life. This is a dead language if we cannot use it in life. We will just forget it. I also teaching computer programming languages like Java in a university. No matter how many examples I have shown the students or how many times I have emphasized the importance of documentation, some students still don't know how to write a simple program, because they are not interested and they don't use it everyday. With any language, if you don't use it, you loose it. My written Chinese has degraded so much because I rarely have time to read Chinese blogs or write Chinese emails because all my friends in China know some English.