Hi, I would like to leverage DataRobot in order to analyze live financial markets and let AI run predictions on the imported data, as well as improve learning on the outcome.
Using an ETH 5m Chart as Database --> Feeding into Datarobot --> Learning --> making upcoming price predictions (for example candle closes) --> feeding back the upcoming result from the prediction into the Datarobot --> Learning again from mispredictions adjusting the data improving --> next prediction
And so on in a loop
Is that possible?
I did pass your questions to my esteem colleagues. Their answer is the question is not if DataRobot can do this, but if it is advisable to do this in your context. Many times understanding the problem is the most important step and then starting with simpler models and then building on top of them.
Here are the questions that you must consider for building a model in finance:
1. how well we understand the factors that contribute to it?
2. how much data is available?
3. how similar the future is to the past?
4. whether the forecasts can affect the thing we are trying to forecast?
I echo my colleague Dalila's points that the fundamental issue of your proposition be addressed before the technical approach be defined. You may wish to take the DataRobot University Time Series Modeling in-person virtual class to further understand some of the considerations required.