Movies View Estimator

This project tries to predict the number of views a movie might have received by considering the movie Rating and the Rating Count.

What makes these project so amazing is that it uses a very small database to train a Linear Regression Model.

It accepts that it is linear data.

The pro for this solution:

• It is easy to understand
• Simple to build
• Very fast to train
• Very small project to maintain

The cons for this solution:

• Too simple
• It might not present a good prediction result
• Small dataset used to train

Another good ways to handle this problem would use k-nearest neighbors algorithm. It is a good fit for the issue.

The pro for this solution:

• It is easy to understand
• Simple to build
• Very small project to maintain

The cons for this solution:

• if the project goes bigger and more data is provided the model is computationally expensive
• Sensitive to noise
• Expensive to predict

How to install:

It uses python 3.8

pip install moviesViewEstimator

How to use:

import pandas as pd
from model import LinearRegressionModel

# if you want to know how many views a movie with rate 8.2 and almost 90 thousand rate count you need to pass a pandas
# dataframe, which is the format accepted by the model
predict_dataframe = pd.DataFrame([[8.2, 89224]]) # rate 8.2 and almost 90 thousand rate count
predict_dataframe.columns = ['Rating', 'Rating Count'] # setting the columns names

model = LinearRegressionModel()
print(model.predict(predict_dataframe))

[2460651]

The predictions tell us that a movie of Rating 8.2 and a Rating Count of almost 90 thousand, would have almost 2.5 million views.

Documentation:

For more information you can check the documentation