What is Regression?
Regression finds correlations between dependent and independent variables. Therefore, regression algorithms help predict continuous variables such as house prices, market trends, weather patterns, oil and gas prices, etc. If your independent variable is time, then you are forecasting future values; otherwise, your model is predicting present but unknown values.
Regression VS Classification
- Regression : Used to predict continuous real values.
- Classification : Used to predict categories or classes.
Pros and cons of regression model
| Regression Model | Pros | Cons |
|---|---|---|
| Linear Regression & Multi Linear Regression | Regardless of the size of the data set. Provides information about the relevance of features. | The Linear Regression Assumptions |
| Polynomial Regression | Regardless of the size of the data set. Works well for non-linear structures. | Choose the correct polynomial degree for a good bias/variance trade-off. |
| Support Vector for Regression(SVR) | Adapts easily. Works well for non-linear structures. Not biased towards outliers. | Need to do feature scaling. Not well known. Hard to understand. |
| Decision Tree Regression | Ideal. No feature scaling required. Works well with linear and non-linear structures. | Poor results on small datasets. Easy to be overfitted. |
| Random Forest Regression | Powerful and accurate. Gives good results on many data sets. | Not ideal. Easy to be overfitted. Number of trees to be determined. |