Handle Outliers - UnivariateHandling outlier is a big task for data scientist. To handle the outliers we have many different methods to handle them i.e. IQR, Z-score, Mean-Median Imputation, Winsorization, etc. We are going to discuss only univariate methods to handle outliers.
Regularization in MLRegularization is used to solve the problem of overfitting caused while training a ML model. In regularization, the model is penalized for overfitting on train data means whenever model tries to predict on training data it add some penalty to the loss function in term of coefficients of the model.
There are about 5 main assumption while training a Linear Regression Model which are:
Relationship of every input feature must be linear with output feature.