NN's and GA's greatly enhance standard trading strategies by making them 'learn' from evolving market conditions. However, these tools do not work in a vacuum. They require inputs that are generally indicative of the behavior of the input. We say 'generally indicative' because NN's and GA's can take mediocre inputs and still come up with good investment decisions. However at AI Investing, we are not satisfied with mediocre inputs and we continually scour the marketplace for good investment signals. These include:

• Simple technical indicators such as moving averages and momentum
• Fundamental information such as earnings and news events
• Wavelets and Spectrum analysis to extract recurring patterns
• Entropy analysis to detect upcoming changes
• Pattern recognition and classification to identify recurring behavior
• Fuzzy Logic to make decision-making more intuitive and flexible