Member-only story
I have been spending the last several blog posts working on how to make predictions on bitcoin because there are so many models that can be used to accomplish this task. I had previously been using the deep learning library,keras tensorflow, to make these predictions, but I thought that it might be a good idea to use the machine learning library, sklearn, to see if I could accomplish the feat that was.
Sklearn is a free open source library made for the Python programming language that provides machine learning functionality to a data science or machine learning project. It provides facilities to carry out supervised, unsupervised, and semi-supervised projects. I personally have spent a lot of time studying sklearn before I decided to expand my sphere of knowledge into the area of tensorflow’s deep learning framework. I think it is a good idea to be fluent in both sklearn and tensorflow because deep learning projects will also have some machine learning aspects attributed to them.
The particular sklearn model that I have used in this project is MLPRegressor. The multilayer perceptron is sklearn’s only neural network, and it comes in two types, being classifier and regressor. In this instance I have used the regressor because the prices are continuous values.
