search

nirajp786 / AlphaInsights

1 stars 1 forks source link

Decision Function #1

Closed Carson-MD closed 3 years ago

Carson-MD commented 3 years ago

Develop the decision function that the agent will use to determine whether or not to make a stock purchase.

For now we will limit ourselves to the data available through the Nasdaq's historical data archives: https://www.nasdaq.com/market-activity/quotes/historical

Attributes include: Date, Closing Price, Opening Price, High, Low, and Volume.

Our initial goal will be to develop a KNN classifier that uses the parameters outlined above to predict whether a stock's price will be higher than it currently is 10 days in the future.

Carson-MD commented 3 years ago

I've created a basic decision function fits a KNN classifier using historical dates and then predicts whether or not the price will be higher in 10 days. Our agent can use this decision function to "buy" when the model predicts that a rise in price will occur.

Note that this model is extremely inaccurate, but its a good starting point to get things up and running.

Carson-MD commented 3 years ago

Prototype decision function was created successfully.