In a dataset of 9 columns: $X_1-X_8, y$.
$y = X_1 * X_5 + X_2 * X_6 + X_3 * X_7 + X_4 * X_8$
Can any form of linear model (anything but SVM, NN, Random Forest, XGBoost, etc.) predict $y$?
$\begingroup$
$\endgroup$
2
-
$\begingroup$ Is this a homework question? $\endgroup$– Lucas MorinCommented Jan 25 at 11:40
-
$\begingroup$ @LucasMorin , No datascience.stackexchange.com/questions/126504/… $\endgroup$– Emad EzzeldinCommented Jan 25 at 12:25
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Only by adding interaction features like multiplication/division/powers..etc. of Features. However, then be sure to cross-validate and test on 0.2 of data size to factor for overfitting.