Recurrent-NeuralNetwork-

The goal of this model is to experiment with how well a basic Recurrent Neural Network with Long-Short-Term-Memory cells can predict the daily adjusted daily closing prices for the company Tesla in December of 2016. First, I retrieved a full year of historical financial data using the Yahoo finance api. Next, I chose which features I would like to use in my model to make a price prediction. Then, I scaled and formatted my data to feed through a Recurrent Neural Network that I programmed using the TensorFlow library in python. Finally, I plotted the network’s predictions and other important indicators to gauge how well my network performed with the financial data I trained it with. Ultimately, my RNN accurately predicted the overall price-direction for a 36 day time period and roughly modeled the daily adjusted close prices to a point where there were distinct similarities between the predictions and actual prices. However, this model was crude at predicting daily price directions."