NASDAQ Example

In the first example, you experimented with predicting functions that can be expressed analytically. On this page, you can use the demo to try predicting financial data - in particular the NASDAQ stock index.

Again, this demo is only for illustration. Predicting the next price is very hard; better results have been reported when predicting supporting indicators, such as a moving average, which can also be useful for trading. Also note that predicting the direction of movement is a completely different task.

How to work with the demonstration

• If you have not seen the first example, please explore it first - the basic description is available there.
• In this version, two data sets are available:
  • NASDAQ daily (i.e., end of day close values) for the whole year 2007 (from January 3 till December 31 of 2007) - 250 values
  • NASDAQ weekly - data from July 5 of 2002 till January 4 of 2008 - 288 values
  For your experiments, it is useful to know that each of these time series behaves as follows: zero for the interval below 0, close values in the interval from 0 to the number of values, and again zero after the last known value. So, for example, to see the whole NASDAQ daily in 2007, set the interval from 0 to 250.
• If you think that the prediction is very good, then try predicting a value farther in the future. When trying to predict a very close value, the network is often just approximating the next value by the last input value, and from a distance the prediction may appear very accurate.