ARTIFICIAL STUPIDITY

As artificial neural nets (ANNs) move from voice and visual pattern re cognition to, the stock market, prices will Hence, analysts need to be affected by their presence. understand their inner mechanics The quic kest way to learn the intricacies of artificial neural nets is through an example of ''the art'' of fitting ANNs. It is an ''art'' because A NNs have both numerous parameters and architectures and there isn't an established methodology for specifying the Optimal architecture The a uthors provide an example of the awesome fitting capabilities of ANN s oftware in the market today, and show how one can achieve the same typ e of results that have made voice and visual pattern recognition softw are become so widespread and successful Finally, while the authors sho w how certain procedures often will not produce excess stock returns, they also show how to apply these procedures to gain excess stock retu rns.