New Register here for RBI, SEBI & NABARD strategy workshop! April 9 at 7 PM

    Question

    Which of the following transformations is most

    appropriate to bring all feature values into the range [0,1] for a machine learning model?
    A Z-score Standardization Correct Answer Incorrect Answer
    B Min-Max Normalization Correct Answer Incorrect Answer
    C One-Hot Encoding Correct Answer Incorrect Answer
    D Logarithmic Transformation Correct Answer Incorrect Answer
    E Ordinal Encoding Correct Answer Incorrect Answer

    Solution

    Explanation: Min-Max Normalization is a technique used to scale features to a fixed range, typically [0,1]. This transformation is particularly useful for algorithms sensitive to the scale of input data, such as gradient descent-based models. This method ensures that each feature contributes proportionately to the model, eliminating bias caused by varying scales across features. Min-Max Normalization is especially suitable for cases where the data has a defined range, making it ideal for neural networks and distance-based algorithms like k-NN. Option A: Z-score Standardization scales data to have a mean of 0 and a standard deviation of 1, which is more suitable for normally distributed data. It does not confine the values to a specific range like [0,1]. Option C: One-Hot Encoding is used for categorical variables, converting them into binary vectors. It is not applicable for scaling numerical data. Option D: Logarithmic Transformation is used to handle skewness in data and is not designed to scale values into a fixed range. Option E: Ordinal Encoding converts categorical data into integers based on their ordinal rank, which is unrelated to numerical feature scaling.

    Practice Next