The repository describes Machine learning methods used to implement a Youtube Video Categorization model which gives a caregory to a video given it's title or description.
- Travel
- Tutorial
- Science
- Arts
- History
- Movies
- Vlog
- Food
- Music
- Fashion
- K-Nearest Neighbours
- Logistic Regression
- Naive Bayes Classifier
- Decision Tree Classifier
- Random Forests Classifier
- Support Vector Classifier
- Ada Boost
- XGBoost Classifer
Algorithm | Hyperparameter | Accuracy |
---|---|---|
Naive Bayes | Alpha - 10 | 86.5 |
Logistic Regression | 0.01, L2 | 88.67 |
KNN | K = 43 | 83.4 |
Random Forests | Tf-IDF | 87.69 |
XGBoost | Lemmatized | 89.97 |
Gradient Boost | Tf-Idf | 88.01 |
SVC | Linear , 'C' = 1 | 88.6 |
I tried the T-SNE algorithm for finding the lower dimensional representation of the data.