Ctrl+K

SVC

SVC#

simple SVC#

        oSvmCls     = SVC(C = C, kernel = 'linear')

    ...
    oSvmCls     = oSvmCls.fit(mXTrain, vYTrain)
    accScore    = oSvmCls.score(mXTest, vYTest)
    ...

kernels#

TBD

LinearSVC#

  • LinearSVC class which optimized SVC with kernel linear as it fits for larger data sets

When using LinearSVC: * If #Samples > #Features -> Set dual = False. * If #Samples < #Features -> Set dual = True (Default).

        oSvmCls     = LinearSVC(C = C, max_iter = maxItr, dual = False)

plot area#

# Grid of the data support
v0       = np.linspace(mX[:, 0].min(), mX[:, 0].max(), numGridPts)
v1       = np.linspace(mX[:, 1].min(), mX[:, 1].max(), numGridPts)
XX0, XX1 = np.meshgrid(v0, v1)
XX       = np.c_[XX0.ravel(), XX1.ravel()]

Z = oSVM.predict(XX)
Z = Z.reshape(XX0.shape)

hF, hA = plt.subplots(figsize = FIG_SIZE_DEF)
hA.contourf(XX0, XX1, Z, colors = CLASS_COLOR, alpha = 0.3, levels = [-0.5, 0.5, 1.5])

option:: probability#

  • in order to have the probability per class on the SVC class we need to set probability = True.

  • probability = True is not always consistent with the decision_function() method. Hence it is better to use it in the case of the SVC

https://scikit-learn.org/stable/modules/svm.html#scores-probabilities

  • enables the estimation of class probabilities for the predictions

  • This option allows you to call the predict_proba and predict_log_proba methods of the SVC - which give you the probability estimates for each class.

Use Cases:

  • Binary and Multiclass Classification: Probability estimates can be particularly useful in binary classification for understanding the likelihood of the positive class. They are also valuable in multiclass scenarios to assess confidence across multiple classes.

  • Threshold Adjustment: In applications where the decision threshold needs to be adjusted away from the default 0.5 (for binary classification), having probability estimates allows for more nuanced control over classification decisions.

  • Risk Assessment: In fields like finance or healthcare, understanding the probability of certain outcomes can be as important as the outcomes themselves for risk assessment and decision-making processes.

  • Ensemble Methods: Probabilities can be used in ensemble learning methods that require them, such as stacking classifiers, where the predictions of several models are combined.

Contents