-
Target : 'normalized-losses'
-
Ignored to feature 'symboling'
-
Using Randomforest regressor model
-
Steps :
- Loading Files
- Data Preprocessing
- fill the missing values
- convert numeric values but object types to float type
- check the distribution of targets
- taking logs to make it normal distribution
- Detecting Outliers Before Modeling
- using box plots to get outliers
- Normalization of numeric features
- One-hot encoding of categorical values
- Modeling(fit & predict & evaluation with MSE)