REPORT OF A PROJECT

ABOUT

PREDICTING THE NUMBERS OF YOUTUBE ADVERTISEMENT VIEWERS

Submitted in fulfilment for the requirement of the award of Internship

in

Machine Learning

Submitted By

KONDRU JAGADISH CHOUDARI

CT UNIVERITY, PUNJAB.INDIA
 ACKNOWLEDGEMENT

My sincerest gratitude goes to my project paper mentor and

Machine Learning coach. It is because of his backing and
support that I was able to complete the report. If there were
mistakes, he helped me correct them by offering various kinds of
assistance accordingly. These are not enough words to state
how grateful I am; lastly but not the least, I appreciate my
beloved parents for always being sources of encouragement and
moral support whose impacts on me were huge in this respect.
Also, I declare that to the best of my knowledge and belief,
Project Work has not been submitted anywhere else

INTRODUCTION
Youtube advertisers pay content creators depending on adviews and clicks for the goods or services they
are promoting The ads can also be estimated using other metrics such as likes, comments among others.
This therefore calls for training a number of regression models so as to select the best model that will
predict adviews which is the problem statement at hand. Data has to be cleaned up before it is fed into
algorithms in order to obtain better results.

Objective

To build a machine learning regression to predict youtube adview count based on other youtube
metrics.

Technology and Concepts

Machine Learning

In classic terms, machine learning is a type of artificial intelligence that enables selflearning from
data and then applies that learning without the need for human intervention.

Linear Regression

Linear Regression is a supervised machine learning algorithm where the predicted output is
continuous and has a constant slope. It's used to predict values within a continuous range, (e.g. sales,
price) rather than trying to classify them into categories (e.g. cat, dog).

There are two main types:

1. Simple regression.
2. Multiple regression

Support Vector Machine

“Support Vector Machine” (SVM) is a supervised machine learning algorithm which can be used for
both classification or regression challenges. However, it is mostly used in classification problems. In the
SVM algorithm, we plot each data item as a point in n-dimensional space (where n is number of features
you have) with the value of each feature being the value of a particular coordinate

Decision Tree
 Decision tree analysis involves making a tree-shaped diagram to chart out a course of action or a
statistical probability analysis. It is used to break down complex problems or branches. Each branch
of the decision tree could be a possible outcome.

Artificial Neural Network(ANN)

An artificial neural network (ANN) is the piece of a computing system designed to simulate the
way the human brain analyzes and processes information. It is the foundation of artificial
intelligence (AI) and solves problems that would prove impossible or difficult by human or
statistical standards. ANNs have self-learning capabilities that enable them to produce better results
as more data becomes available.

Data Description

The file train.csv contains metrics and other details of about 15000 youtube videos. The metrics
include number of views, likes, dislikes, comments and apart from that published date, duration and
category are also included. The train.csv file also contains the metric number of adviews which is
our target variable for prediction.

Steps For adview prediction

1. Import the datasets and libraries, check shape and datatype.

2. Visualise the dataset using plotting using heatmaps and plots. You can study data distributions
for each attribute as well.

3. Clean the dataset by removing missing values and other things.

4. Transform attributes into numerical values and other necessary transformations

5. Normalise your data and split the data into training, validation and test set in the appropriate
ratio.

6. Use linear regression, Support Vector Regressor for training and get errors.

7. Use Decision Tree Regressor and Random Forest Regressors.

8. Build an artificial neural network and train it with different layers and hyperparameters.
 Experiment a little. Use keras.

9. Pick the best model based on error as well as generalisation.

10. Take the test dataset test.csv

11. Clean the test dataset by removing missing values

12. Remove unnecessary columns which has no impact to target variable

13. Transform the categorical attribute to numerical attribute.

14. Find prediction using the best algorithm

15. Save it into a new csv file by naming as Predictions_Submission.csv

Visualization :

This is the histogram of “Category” column

This is the histogram of “adview” column

This is the heatmap which shows the co-relation of all columns with each other.
Table:

Algorithm Linear Random Decision tree Support ANN

Regression forest vector
machine
Mean 3707.37800 3274.69029 3059.31079 3707.37800 3304.26489
Absolute 5824529 66905504 2349727 5824529 4606637
Error

Mean 835663131. 644433788. 1226286165 835663131. 829552666.

Squared 1210335 0361483 .4118853 1210335 7955565

Error
Root Mean 28907.8385 25385.7004 35018.3689 28907.8385 28801.9559
Squ 7573986 6376795 713254 7573986 5433679
ared Error

Best Model
From the training dataset by applying all algorithms for train the model,we found that "Random
Forest Regressor" algorithm has less root mean squared error as compared to othre algorithms.As
we know model having less root mean squared error is more perfect.So here for prediction of test
dataset we use "Random Forest” algorithm.

Conclusions

We had a lot of different ideas for the project, but were maybe originally too ambitious for our
goals. We were originally trying to predict the view count of advertisement. In this way we can
predict the adview of an advertisement. We were hoping that. Some more things that we could have
tried if we had more time would include.