I am working on some new project In which I am using using machine learning algorithm. Some of the recent report on Covid-19 showed that the predictions or projected statistics using an Machine learning algorithm did not match with the actual value ? What could be the reason for this ? what is the method to ensure that ML predictions are more accurate to the actual values ?
Machine learning is workable if the underlying pattern is known or can be traced. For Covid-19, this is not yet known
Machine learning algorithms need a large amount of data in the training phase in order to get high accuracy. Moreover the vector data should contain all possible prediction depends data, things that we don't cover to this moment. But we can create a model or a pattern that converge to the reality (time infinity) and that have a specific error, torelated or not. But in my opinion : having an information with 0.3 true probability more than nothing.
You can check my paper Article A Smart Architecture Design for Health Remote monitoring Sys...
- Badges
- Science topic