GEN: Neural Net Inputs for Futures Index Forecasting

[Peter Lim <peterl@xxxxxxxxx>]

Hi,

As a beginner in the use of neural net forecasting of the movement of index
futures e.g. S&P, DJIA futures and the australian SPI , I understand that
the use of Open, High, Low, close, volume data is not desirable as they
represents non stationary items and the use of smoothed stationary items
such as technical indicators preprocessed eg. Stochastics, RSI, MACD,
momentum are better.

>From personal experience, what would be desirable inputs for forecasting
the S&P futures or DJIA futures? Would anyone experienced in use of neural
nets care to share? Thanks.

Peter Lim