Some studies used first preprocessing, feature selection, data balaning, model building etc. and some follows pre-processing, data balacing, feature selection etc.
Can we adopt sequence which provide better result or is there mandatory to follow any sequence?
There are three types in Multiple Linear Regression (MLR). Standardized (researcher enter all the IV at once), sequential, (forward, backward, where the researcher leaver for the statistical program the freedom to enter the IV based on a preset criteria, which mostly alpha level), the Hierarchical regression (through which research can control the order of variable entry in every step based on a theoretical reason/ chronological order of variables occurrence; in addition, this type is used to control for variables in early steps)
- Badges
- Science topic