Q3) [ Marks: 5 + 5 = 10 ]

Consider you want to develop a classification model using the machine learning method. You observed that the fluctuation of input variables is quite high and the data set has many outlier observations. Which type of machine learning method is suitable in such a situation? Provide sufficient explanations of your answer.

Consider another situation where you want to develop a prediction model on numerical output data and your main aim is to improve the predictive accuracy with respect to target value not variance reduction. Which machine learning model is useful for this particular situation? Include sufficient explanations to justify your model selection.