Instruction: Detail your approach, including data collection, preprocessing, and model selection.
Context: This question tests the candidate's understanding of financial market predictions using machine learning, with a focus on data handling and model architecture.
Official answer available
Preview the opening of the answer, then unlock the full walkthrough.
I would start by reframing the problem slightly, because directly predicting stock prices is usually less useful and less stable than predicting a return, direction, volatility band, or some other decision-relevant target over a defined horizon. The target definition should match the trading or risk decision the model is meant to support.
For data,...