The matrix stick-breaking process for flexible multi-task learning

Computer Science

• Multitask Learning 100%
• Dirichlet Process 66%
• Classification Models 33%
• Sampling Algorithm 33%
• Clustering Task 33%
• Share Information 33%
• Feature Vector 33%

Keyphrases

• Matrix Stick-breaking Process 100%
• Task Sharing 33%
• Feature Component 33%
• Task Clustering 33%
• Local Clustering 33%