Скачать или смотреть Can Convolutional Neural Networks Predict Stock Prices?

Follow this link to read my written articles on applying data science to investing: https://www.eddywealth.com/articles/?...

Github repository containing the code: https://github.com/moneygeek/cnn-stoc...

00:00 Introduction
00:10 Concepts
02:28 Code
11:23 Results

Convolutional neural networks (CNNs) are one of the most popular machine learning architectures, used most famously for image recognition. Can CNNs be used to predict stock prices? I offer an answer.

This video consists of 3 chapters.

In the first, I discuss why CNNs' mechanisms could be well suited to determine the direction of stocks. I explain how CNNs work for image processing, and draw analogues with how technical analysts approach trading.

In the second chapter, I walk through the code that implements a CNN model, paying attention to only the meaningful parts of the code. I explain some sections of the code using graphics.

In the third chapter, I discuss how well the model is able to make forecasts. I also discuss ways to improve the model.

Disclaimer: This video is for educational purposes only. It doesn't constitute advice.