Fuzzy Candlestick Patterns Neural Network Algorithm

25 January 2018, 16:24
Ahmad Hassam

Candlestick patterns are leading signals for trend reversal and trend continuation.

However interpreting these candlestick patterns is an art that requires years of experience.

Candlestick patterns are most of the time vague and imprecise.

Can we use these candlestick patterns in machine learning and data mining.

I have written a post in which I explain in detail a Fuzzy Candlestick Pattern Algorithm.

I use a simple Multi Layer Perceptron to build the model and achieve 82% accuracy.

By using deep learning, we can take the accuracy of the model above 90%.

Fuzzy logic is the key  behind this successful model.

Many traders have no idea what is fuzzy logic and how to use it in trading.

In our life we find everything vague and imprecise just like the candlestick patterns.

Nothing is black and white like in the tradition logic system where everything is either yes or no.

Fuzzy logic solves this problem by making everything grey like that in our life.

Fuzzy logic is ideally suited for developing algorithmic trading models.

Share it with friends: