This project was done as a part of my B.Tech project.
The intial aim was to study different Machine learning algorithms to automatically tag youtube videos into different game categories, but later, the focus was shifted from this to :
Finding different methods of fusion of the data, to create a better machine learning model.
We used spectral clustering to find better ways of fusion. We group the correlated features into some clusters and then, apply the classifier algorithm on each of these clusters, and finally combine the result from each classifier.
Dataset used : YouTube Multiview Video Games Dataset
Link to the dataset :https://archive.ics.uci.edu/ml/datasets/YouTube+Multiview+Video+Games+Dataset
Research Papers followed : http://link.springer.com/article/10.1007%2Fs10994-013-5377-0/fulltext.html
Right now.