Your company manages a video sharing website where users can watch and upload videos.
You need to create an ML model to predict which newly uploaded videos will be the most popular so that those videos can be prioritized on your company's website.
Which result should you use to determine whether the model is successful?
Click on the arrows to vote for the correct answer
A. B. C. D.C.
The correct answer for this question is option C: The model predicts 95% of the most popular videos measured by watch time within 30 days of being uploaded.
Explanation: The goal of the ML model is to predict which newly uploaded videos will be the most popular, so that they can be prioritized on the company's website. Therefore, the success of the model should be measured based on how well it is able to predict popular videos.
Option A, "The model predicts videos as popular if the user who uploads them has over 10,000 likes," is not a good measure of success. This is because the number of likes a user has does not necessarily correlate with the popularity of the videos they upload. It is possible that a user with a lot of likes may upload a video that is not popular, or that a user with few likes may upload a very popular video.
Option B, "The model predicts 97.5% of the most popular clickbait videos measured by number of clicks," is also not a good measure of success. This is because the goal is not to predict clickbait videos, but rather to predict popular videos. Clickbait videos may receive a lot of clicks, but they may not necessarily be popular or have high engagement.
Option C, "The model predicts 95% of the most popular videos measured by watch time within 30 days of being uploaded," is a good measure of success. This is because watch time is a good measure of engagement and popularity. If a video has a high watch time, it means that viewers are interested in the content and are spending time watching it. Additionally, the 30-day timeframe is a reasonable measure of success, as it allows for a sufficient amount of time for the video to become popular.
Option D, "The Pearson correlation coefficient between the log-transformed number of views after 7 days and 30 days after publication is equal to 0," is not a good measure of success. This is because a correlation coefficient of 0 means that there is no relationship between the number of views after 7 days and 30 days after publication. However, this does not necessarily mean that the model is successful at predicting popular videos. Additionally, the measure of success should be based on the ability of the model to predict popular videos, not on the correlation coefficient between two variables.