[Hassam]

Deep learning is very powerful. In the recent years deep learning has solved many artificial intelligence problems. Tensorflow is a open source deep learning framework developed by Google. Developing deep learning model can be easy. This is what most books says. Deep learning is supposed to read the data and find the features that can make the best predictions. But in reality we have to select the features and then use the deep learning model. Below I have made a video where I train a deep learning neural network model using Tensorflow and try to make prediction about the next candle. Now this is just a toy model I don't think we cannot use it in live trading. But as a first prototype we can play with this deep learning neural network model. In the future I will post more better deep learning neural network models. Feature selection is very important when building a deep learning neural network model. Financial markets are hard to predict. The main reason is non stationary nature of financial time series. Financial markets are sentiment driven now a days in short run. We have so much breaking news. Each breaking news has the potential to move the financial market a lot. So building deep learning neural network models for predicting financial markets is not an easy task. The important question is can a discretionary trader beat the market more or a deep learning trading system?