I'm trying to forecast some values with information about many variables from the past, there's a description of the data:

It contains information over the time from 5 sensors for different variables ie:

sensor var1 var2 var3 target date
sensor1 ... ... ... ... ...
sensor2 ... ... ... ... ...

and so on... after a transformation data becomes

sensor var1 var2 var3 var1(t-1) var2(t-1) var3(t-1) target(t-1)
sensor1 ... ... ... ... ...
sensor2 ... ... ... ... ...

Until t-8.

However, when forecasting using sklearn linear regressor I see that there's a delay and the model is almost telling that the next value is the previous one. Have you faced this before? what you've done? is this wrong? :( Please help

Thank you!

  • $\begingroup$ what software is this? $\endgroup$ – user54285 Mar 24 at 22:59
  • $\begingroup$ I didn't understand D: I'm using python, pandas and sklearn $\endgroup$ – CrazyMath65 Mar 25 at 2:51
  • $\begingroup$ Sorry I know nothing of python, thus my question. I was not really sure what you were asking. I use R and SAS. $\endgroup$ – user54285 Mar 25 at 20:57
  • $\begingroup$ No worries, what I'm having currently is a lag between the actual and forecasted value. I'm using a lag of 15 days for the target variable and other variables and I wanted to knwo if there's something wrong and how to fix it :(. $\endgroup$ – CrazyMath65 Mar 25 at 21:54
  • $\begingroup$ You can use lags to predict anything. There are dangers including autocorrelation, non-stationarity, and many other factors which are serious. Since I don't know python how exactly (what method not tool) are you using to predict. There are many types of time series and its hard to answer not knowing which approach you are using. I am not truly an expert in any case, I just read time series. $\endgroup$ – user54285 Mar 25 at 23:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.