Predicting Elliott Flat and Zigzag Internal Shapes by Statistical Learning on Fibonacci Ratios

Elliott waves and Fibonacci proportions can be used to estimate the price behavior of an asset since they can describe the patterns and relationships from time series of an asset historical price. The challenge is projecting future patterns from a sequence of patterns already mapped from historical data. This paper presents a way to predict the internal shape of the Flat and the Zigzag patterns that happen in Elliott waves. The results show that our model was able to reduce the error 4 times when compared to a solution that is guessing the length only by respecting Elliott wave rules.