Hi Gurus,
I found the knime did support single variable or single feature “Linear Regression Learner”. This function is limited productive at Enterprise environment.
So, my question is:
Whether does knime support multi-features machine learning?
Could you give us an example what you mean by “multi-features machine learning” then we might be able to point you towards model building options within KNIME for yes/no, multi-target or regression models. Or maybe something else if necessary.
ok. thanks. I will submit my real business scenario. I need build the model from the following data points:
tagTime (datetime), ground temperature(float), supplying heat water temperature(float)
supplying heat water temperature is our linear regression target. But I can only select ground temperature as feature, not tagTime.
You can convert a date and time into a relative number like “Day of year” and the time of day as a numeric or time variable. Also you might extract features like the number of the month or the week since that might play a role. I am not sure if you will get results just with a regression but that is how you might use the Date and Time information.
Make sure you can reuse it the following year/time period, so no year 2018 or something since that will not be present in the future.
You might think about if some time series model might also help you, but that depends on your goal.
1 Like