Timeline for Random forest regressor accuracy reduces when the input data is not shuffled
Current License: CC BY-SA 4.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 24, 2022 at 22:05 | comment | added | lsr729 | Thanks for the suggestion, I have put the question in a detailed manner here- stats.stackexchange.com/q/569020/275463. | |
| Mar 24, 2022 at 22:04 | vote | accept | lsr729 | ||
| Mar 26, 2022 at 18:36 | |||||
| Mar 24, 2022 at 21:56 | comment | added | Sycorax♦ | Random forest assumes that the data are iid realizations from some data-generating distribution, but your data are arranged in time. You should split your data based on time, as described in the linked threads, and you should use a time-series model. | |
| Mar 24, 2022 at 21:54 | comment | added | lsr729 | I see. What other model can you like to suggest in this problem? | |
| Mar 24, 2022 at 21:50 | comment | added | Sycorax♦ | It doesn't know about time, it just knows about things that are correlated with time, and that they're also correlated with the outcome. Random forest particularly is bad at extrapolating outside the range of the training data, so a strong trend in your target means it will be hard for the model to predict the future, but much easier to predict the past if you tell it the future. | |
| Mar 24, 2022 at 21:34 | comment | added | lsr729 | That's true, but how does the model know that the features depend on time? I believe the model should look for only the independent data provided to it in the training. Is it probably because the response variable decreases with time (I have uploaded the plot in the question)? | |
| Mar 24, 2022 at 21:13 | comment | added | Sycorax♦ | Sure, but even without timestamps, the features vary with time, and knowing the future means you can predict the "gaps" in the time-series more precisely. | |
| Mar 24, 2022 at 20:44 | comment | added | lsr729 |
Thanks for the answer and links. I have a question- what's probably happened is that shuffling the data is letting the model use information from the future to predict the past why does the model care about the order of the data? I am not including the time as independent variable, I am just feeding the model some shuffled data.
|
|
| Mar 24, 2022 at 18:27 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 638 characters in body
|
| Mar 24, 2022 at 8:57 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 44 characters in body
|
| Mar 24, 2022 at 8:48 | history | edited | Sycorax♦ | CC BY-SA 4.0 |
added 19 characters in body
|
| Mar 24, 2022 at 8:37 | history | answered | Sycorax♦ | CC BY-SA 4.0 |