Technical Indicator Engineering for Machine Learning: Volume I: Detecting Major Stock Market Bottoms with Moving-Average Diffusion Indicators

This work describes the NV-PA-PV Sequence, a method for detecting major stock market bottoms using moving-average diffusion indicators. It introduces a filtering and weighting method to enhance the signal to noise ratio of oscillators. We show the performance of buy signals back to 1962 using the S&P 500 Index. We show how to create an ensemble of alarm indicators derived from moving average diffusion indicators. These indicators should be useful to both practitioners of machine learning and discretionary traders.