Thanks for sharing. Nice video. This is interesting, but it looks a little limited. Is it really just analyzing a single variable and predicting the pattern based on it? Can it take multiple inputs and predict the future? Say S&P 500, PE-ratios, and federal interest rate to predict the future of S&P 500?
Great question, and totally agree on the limitations of MV forecasting. This is what one of the engineers behind it has to say... '@harshitv804 as we discussed in the paper, Chronos currently focuses on univariate forecasting. For multivariate time series, you might want to use Chronos on the individual dimensions independently. If you have specific multivariate use cases/datasets to share with us, please do. It will helpful for us to understand the types of practical multivariate problems.' Source: github.com/amazon-science/chronos-forecasting/issues/13
@@CalifornianViking I assume not. Because the way it works is that you have historic datapoints for a single metric or measure, e.g. the S&P 500 (so lets say the weekly value for the past 3 years). Then you use that historic set of values, to then predict the S&P 500 for the next 6 weeks (could be shorter could be longer). What you can do though is run different forecasts for each metric and possibly overlay/ sense check them against each other.