So, this newsletter will occasionally be intermittent -- at least until circumstances in my busy life arrange themselves otherwise.  If this project turns out to be of little value to either myself or others, it will probably die a quiet death.  If, on the other hand, the insights and forecasts provided herein prove to be of sufficient merit to put this project on a businesslike footing, then it will continue.  This current test will run for ninety days. 

Over time, this letter will cover a variety of forecasting approaches: neural networks, technical analysis, fundamental analysis, associative remote viewing and other intuitive approaches. 

To begin with, the neural network forecast posted below was created using a program called Profit written by BioComp Systems. I have been working extensively with this program since March of 1999. And I regard it as one of the most effective tools available for end-of-day market forecasting. The results shown below are produced by a "mega-system" that contains the votes of over a dozen unique, neural network systems. Each of these contains the votes of over 100 neural network models. I believe that these forecasts are about 60% accurate, overall. 

Each day, I generate a forecast for the following two days. 

More will follow in the days and weeks to come. 

Jeffrey 
 

Neural Network S&P 500 Index Forecast: 
Monday: negative 
Tuesday: negative 

Return to ForecastingSystems.com