Today's learning :-
1. Feature selection - There are two way to select feature - Manual process and using a Program
2. Feature selection using Program - We have three algorithms for this (Filter, wrapper and embedded(Lasso/L1 regularization) )
3. Feature selection using Filter method-
4. We have multiple way for feature selection using filter method like -
(A) Correlation(pandas.corr) - Here we have to check which column is highly correlated and which one is partial or less related(0.25, 0.33, 0.45 etc)
(B) Constant Variables - In this we just check constant columns and eliminate them. For this we use VariableThreshold(threshold=correlation %)
5. Multi-Linear Regressions. If Y(Target is depended on more than a predictor X (X1, X2, X3, ...Xn)) then multi-linear comes into the picture.
6. Written our own code to use webcam for live streaming using CV2
1. Feature selection - There are two way to select feature - Manual process and using a Program
2. Feature selection using Program - We have three algorithms for this (Filter, wrapper and embedded(Lasso/L1 regularization) )
3. Feature selection using Filter method-
4. We have multiple way for feature selection using filter method like -
(A) Correlation(pandas.corr) - Here we have to check which column is highly correlated and which one is partial or less related(0.25, 0.33, 0.45 etc)
(B) Constant Variables - In this we just check constant columns and eliminate them. For this we use VariableThreshold(threshold=correlation %)
5. Multi-Linear Regressions. If Y(Target is depended on more than a predictor X (X1, X2, X3, ...Xn)) then multi-linear comes into the picture.
6. Written our own code to use webcam for live streaming using CV2
Comments
Post a Comment
Please share your experience.....