Question 2 - Manipulating data [25 marks] (a) Two datasets are given below. Answer the following questions, and give numerical answers in 2 decimal places. [7] Dataset A: (1, 3, 3, 5, 9) Dataset B: (1, 3, 3, 5, 9,99) (i) Calculate the mean, median, standard deviation of dataset A. (ii) Calculate the mean, median, standard deviation of dataset B. (iii) The difference between the two datasets is the additional outlier value 99 in dataset B. Which of the mean and the median does the existence of the outlier have a more severe impact on? (b) A dataset is given below. Answer the following questions in 2 decimal places. [6] Dataset A: (1, 3, 3, 5, 9) (i) Normalize the dataset to the range -1.0 to 1.0. (ii) Standardize the dataset to the mean 0.0 and the standard deviation 1.0. (c) Apply binning to the values (7, 1, 12, -6, 6, 5, 14, -1, 16, 4, 10, 7) using bins of size 3 and the means of the bins. Do it manually without programming; show the steps, the contents and replacement of the bins. [6] [6] (d) Two jobs of cleaning data are smoothing data and handling missing data. (i) Smoothing data, such as binning, typically discards some information from the data. Explain why smoothing data is useful in machine learning. (ii) Name two techniques of smoothing data other than binning. (iii) Two ways of handling missing data are removing records with missing values and filling in missing values. Suggest the criteria or condition of selecting between the two ways with reference to the missing values and the data.