Add LSTM prediction utility function(s)

[jaARke]

Once we have decided on an LSTM model configuration, we should add support for making predictions on a specific stock. Specifically, we should add function(s) in utils/quant/ that accept a stock ticker, a data collection timeframe, and a prediction timeframe as parameters and returns a price prediction for each day in the prediction timeframe.

The sequence of events inside these function(s) is as follows:

• Receive historic data about the queried stock price over the given collection timeframe (see this issue).
• Cleanse / transform the data to fit our needs.
• Train the model on the stock's price data.
• Make predictions based on the prediction timeframe.

In addition to the daily predictions, it might be nice to return any performance metrics available after the analysis.