Say you’re an ML engineer for a large bank. Your department already has several well-functioning forecasting models for stock price predictions. These models take in streams of data regarding stocks to predict future stock prices.
However, with any predictive model, there exists a certain degree of uncertainty. Your manager asks you to quantify the uncertainty of a current model so she has a good idea of expected success in the future.
Given a time-series forecasting model and data about past predicted and true stock prices, how would you quantify the uncertainty of the current model?Next question: Multi-Reaction