Estimating when a video gets 1K views using Python & Scikit learn
================================================
Source code: https://github.com/maksimKorzh/ml-tut...
================================================
================================================
VICI - Minimalist cross-platform terminal based text editor
================================================
https://github.com/maksimKorzh/vici
=====================
SUPPORT THE CHANNEL
=====================
Patreon: / code_monkey_king
PayPal: [email protected]
==========
CONTENTS
==========
00:00 Intro
00:50 Exploring dataset
02:14 Coding starts here
02:30 Importing required packages
04:35 Creating virtual environment
05:15 Installing required modules
06:18 Loading data from CSV
06:56 Categorical features explained
07:45 When regression problem turns into classification problem
08:55 Dealing with random noise spoiling the model
09:15 How feature engineering is going to be accomplished
10:04 Preparing dataset
10:30 Extracting programming languages from title
15:15 Adding programming language feature to dataset
17:00 Extracting video topic from title
21:25 Adding video topic feature to dataset
23:23 Using views as a model prediction target
24:25 Adding prediction target to dataset
25:20 Converting Y/M/W to days
27:55 Adding days to dataset
29:18 Filtering dataset
30:50 Inspecting filtered dataset
33:00 Splitting data into input features and target variable
35:05 Splitting data into training and testing sets
37:15 Applying Decision Tree Classifier algorithm to train the model
38:50 Making predictions and measuring model accuracy
39:37 TESTING MODEL
40:12 Improving model accuracy by adding extra feature
44:17 Writing a script to use an existing model to make predictions
45:20 Loading pre-trained model
46:40 Creating data frame to encode user input to
47:15 Encoding features
52:13 Estimating when this video is going to get 1000 views
54:20 How did I obtain the dataset