2012年4月10日 星期二

What Is Dynamic Data?

"Data" in the field of computer science is a term that simply refers to information that is in an electronic form and that usually is stored in a database. "Dynamic" denotes change, and when the word is used to describe data — as in "dynamic data" — it refers to electronic information that changes as needed or desired. There are many reasons why data would need to be dynamic. For example, large electronic commerce web sites that offer many different products for sale almost always have to keep track of inventory. Product information is stored in a database and is retrieved and updated in real time.

If one visitor purchases the last of a product, a "sold out" notice might be programmed to display to subsequent visitors. The electronic information displayed would reflect the changes in the database that were made as a result of the purchasing activities of cyberspace customers. This is a prime example of dynamic data in the real world.

Most websites are database driven or simply data driven. This means that the content of such sites is created on the fly, based on ever-changing conditions. For example, typing in the web address of a site and landing on the home page is a condition for which a web developer might program the display of dynamic data to occur.

The developer might want the content of the home page to be displayed in a random order so that the page appears different each time it is called. He or she might also might want to do something like displaying the most recent content added to the database or even allowing visitors to customize, to some extent, their visitor experience. Personal preferences of each visitor would determine what content is displayed and how that content appears to them. This is another real-world example of the generation of dynamic data.

Web developers use programming languages for the coding of scripts to achieve dynamic data. If a website is data driven, the developer will have to work with a database. The developer can write queries to add, update, delete and combine information in the database to create dynamic data that will be displayed for all possible scenarios.

A database is not always what is behind data that is dynamic. The time changes from minute to minute, and the date changes each day. Simple scripts can be written to achieve the display of the current time and date on a website that consists of static or unchanging data.


2011年6月20日 星期一

Random walk和非線性模型

Karl Pearson(1905)在Nature雜誌上提問:假如有個醉漢,醉得非常嚴重,完全喪失方向感,把他放在荒郊野外一段時間後,再去找他,在哪裡找到他的機率最大?

Random walk是指無法預測下一期的變化,亦即未來是會隨機漫步。如果時間序列具有Random walk的性質,則為非定態時間序列。 非定態時間序列在模型的估計上會發生Granger and Newbold(1974)提出的假性迴歸(spurious regression)問題。而要檢測變數是定態或非定態,最常使用的方法為單根檢定。一般常見的單根檢定有ADF、PP、KPSS

Granger and Terasvrita (1993)也發現大多數的總體經濟變數大多呈現非線性的趨勢。倘若變數間具有非線性關係,則在利用線性模型來進行分析,可能會發生一些specification error的問題。為了得到更精確的實證結論,因此必須考慮非線性模型的應用。現在常使用的非線性單根檢定有KSS單根檢定及LNV單根檢定。

2011年3月17日 星期四

Research Topics

1. Service-Oriented Dynamic Data Driven Application Systems

2. Service Selection

3. Service Composition

4. Multi-Agent Simulation for Services Design

5. Patent Prior Art Retrieval