Make a ClassificationPredictor class for seamless predictions from various models

[abhi8893]

class ClassificationPredictor:

    def __init__(self, model, class_mode, class_names=None, featurizer=None):
        self.model = model
        self.featurizer = featurizer
        self.class_mode = class_mode

    def predict(self, X, featurize=False):
        pass

    def predict_proba(self, X, featurize=False):
        pass

Features:

• Add a featurizer option
• Model can be a simple predict proba function, or a sklearn model, or a tf.keras model (Yet to decide how to handle this)