4 ms·
Can you provide any links to sites that take current data and use better models to forecast? This site strikes me as a useful first step. But clearly there are
by skmurphy 7y ago
Can you provide any links to sites that take current data and use better models to forecast? This site strikes me as a useful first step. But clearly there are better approaches, I have only seen predictions in papers, not in an equivalent site.
- sgt101 7y agoThere are a few around. Try this one https://www.buyupside.com/calculators/sir.php https://www.buyupside.com/calculators/sir.php
- skmurphy 7y agoI think the current forecast is a reasonable first approximation for a three day forecast. The model you propose is more accurate but requires many more parameters, each with their own error bands. I think you are looking at a very complex rapidly evolving situation characterized by an eventual transition between exponential growth and S-curve transition that is very hard to predict. The chart now indicates the forecast points and assumes 3 days of exponential growth, which until the virus is contained is a reasonable three day forecast. I have no affiliation with the author of the chart but appreciate the thought and effort that went into it.