Creating Your Record
Using only 8 easily accessible attributes predict the closing price of the top 2000 cryptocurrency prices. The dataset used for training can be viewed in the attached zip file in this solutions document. There are eight different attributes which must be used when attempting to create the CSV record for making your prediction. The eight different attributes used for predicting the closing price of a cryptocurrency are described below.
- Rank: The current rank of the cryptocurrency in the market. (1-2000)
- Open: The opening price of the cryptocurrency in $USD.
- Yesterday's High: Yesterday's highest price of the cryptocurrency in $USD.
- Yesterday's Low: Yesterday's lowest price of the cryptocurrency in $USD.
- Yesterday's Volume: Yesterday's volume of the cryptocurrency traded in the market $USD.
- Market: Yesterday's market cap in $USD for the cryptocurrency.
- Yesterday's Close Ratio: Yesterday's (Close-Low)/(High-Low).
- Yesterday's Spread: The $USD difference between the high and low values for the day.
Example sample data taken from Bitcoin:
Making your Prediction
When making your predictions with SageMaker the model will output a decimal number that represents that predicted closing price for the day.