Efficient and High-quality Recommendations via Momentum-incorporated Parallel Stochastic Gradient Descent-Based Learning

Keyphrases

• Parallel Stochastic Gradient Descent 100%
• Stochastic Gradient Algorithm 75%
• Convergence Rate 50%
• Latent Factor Model 50%
• Main Idea 25%
• Computational Efficiency 25%
• Existing State 25%
• Training Process 25%
• Learning Algorithm 25%
• Industrial Problems 25%
• Parallelization 25%
• Low Efficiency 25%
• Recommender Systems 25%
• Sparse Matrices 25%
• Momentum Effect 25%
• Splitting Strategy 25%
• Stochastic Gradient Descent 25%
• High-dimensional Matrices 25%
• Computational Scalability 25%
• Industrial Recommender System 25%
• Latent Factor Analysis 25%
• Serial Mechanism 25%
• Data Splitting 25%

Computer Science

• Gradient Descent 100%
• Recommender Systems 33%
• Convergence Rate 16%
• Learning Algorithm 16%
• Experimental Result 16%
• Training Process 16%
• Parallelization 16%
• Computational Efficiency 16%

Chemical Engineering

• Scalability 100%