new to MS SQL Server Analytics - Stock Price Trend Analysis

Hi,

I am new to Data Mining and have begun using SQL Server 2005 Analytics.

I want to identify a sudden increase/decrease [high volatility] in stock prices. My DB Table consists of Attributes:

1) Stock Price at end of trading day
2) Date [Data is collected for each stock once every trading day]
3) Daily forward return for stock = (Today's Price - Yesterday's Price)/Today's price
4) Name of the Stock for which above three values were collected [ie, Ticker Name]

I am not able to understand which Mining model will help me get a list of high volatility days in stock prices for each stock. [All I need is Name of Ticker and Date of High Volatility]

Any advice?

Thank you,
Tarun Makhija

I believe this question was answered in the newsgroups, reposting the answer here:

"I recommend constructing a bunch of variables such as

- large burst 1 day ago
- number of medium bursts 1-2 weeks ago
etc.

You can then predict FWR from these inputs using a nerual net.

David "