YOUTUBE ALGORITHM MODELING WITH NEURAL NETWORKS FOR VIDEO PERFORMANCE ANALYSIS
With the shift from the conventional world to the virtual world, understanding Internet algorithms is essential to adapt to global developments, particularly in the realms of information and commerce. This research involves the development of a machine learning model capable of predicting the vie...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/81627 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | With the shift from the conventional world to the virtual world, understanding
Internet algorithms is essential to adapt to global developments, particularly in the
realms of information and commerce. This research involves the development of a
machine learning model capable of predicting the view count of a Youtube video
by analyzing the weight of each factor influencing the video's performance, namely
title interestingness, video interestingness, video retention, number of subscribers,
and video response. The model was developed using a neural network algorithm,
resulting in a model that can predict the number of viewers of a video with an error
rate of 8.62%. Additionally, the model reveals that the most influential factor on the
number of views in the first layer is the number of subscribers a channel has, with
a weight percentage of 72.03%, and video quality in the second layer, with a weight
percentage of 84.33%.
.
|
|---|