Shani has been asked to come up with a scoring system for crowd-sourced online translators (i.e. people who contribute some of their spare time to translating sections of online documents into different languages). The data she has access to are:

1) Quality (An average score out of 10 based on whether online users approve of the translator's work).

2) Time (the average time it has taken for the translator to complete an assigned task - ranges from 3 days to 10 days).

3) Productivity (indicates how many projects the translator has completed in the past month. Ranges from 8 to 50).

Which of the following are appropriate data transformations that could be used after she collects data relating to these three features? After pre-processing the data she wants all three criteria to have a score between 0 and 1 so that she can give an overall rating capturing the quality and efficiency of the translators. [Multiple answers could be selected]

Options in Screenshot

| C) Linear feature scaling (without any negation) can be used on the quality variable, i.e. of the form f(t) : 1t0, so that the data ranges from 0 to 1. A negation function, f(t) = 13 - t, can be used for the time variable, after which it could be scaled so that it ranges from 0 to 1. A decreasing function, N(t) = 58 - t, can be used for the productivity variable, after which it could be scaled so that it ranges from 0 to 1. (3 Linear feature scaling (without any negation) can be used on the time variable, i.e. of the form t : 753 , so that the data ranges from 0 to 1. 103 C) Linear feature scaling (without any negation) can be used on the productivity variable, i.e. of the form f(t) : 5:50:88 , so that the data ranges from O to 1. C] A negation, N(t) = 10 - t, can be used for the quality variable, after which it could be scaled so that it ranges from 0 to 1