Accelerating Recommender Systems using GPUs
arXiv:1511.02433v112 citations
Originality Synthesis-oriented
AI Analysis
This work provides faster processing for recommender systems, but it is incremental as it applies existing methods to new hardware.
The paper tackled the problem of accelerating recommender systems by implementing GPU versions of CCD++ and ALS algorithms, achieving a maximum speedup of 14.8 compared to multi-core versions.
We describe GPU implementations of the matrix recommender algorithms CCD++ and ALS. We compare the processing time and predictive ability of the GPU implementations with existing multi-core versions of the same algorithms. Results on the GPU are better than the results of the multi-core versions (maximum speedup of 14.8).