5.1 SVM Flashcards
Whats the meaning of 0-weighted linear weights in SVM?
What that means with zero classifier coefficients is that there is probably an equivalent or better predictor of your data present in the other features (i.e., it’s redundant). It may not be an intuitively obvious choice, but it’s a choice that that particular algorithm found. A different algorithm (e.g., LibLinear) may have something different to say.
Which is the best hyperplane for the dataset in the figure?
Hyperplane b maximises the margin between the two classes.
Which of the following descriptions are true about ‘soft margin’ for SVM?
Soft margin is applied to trade off between maximising margin and minimising training errors. The magnitude of errors is determined by the hyperparameter C and the slack variables of the optimisation result.
Soft margin is an extension of hard margin SVM where the data is not linearly separable.
What does the parameter C stands for in SVM?
The choice of C affects how much training error is allowed by SVM. A large value of C leads to a smaller margin given that hyperplane minimises the training errors. A small C parameter looks for a larger margin which will result in more misclassifications.
The performance of SVM depends on:
ALL options
SVM is a supervised ML model which is applicable to both linear and non-linear classifications. The application of SVM covers many aspects of real-world problems. The above are only a few examples.
Which of the following are suitable for SVM classifier?
