Open Dikosh opened 5 years ago
Hi. This is a good idea. I guess you want this to scale up the number of predictions you can make?
yes
You could save the Recommender object and then run several instances of R and load the Recommender object to create recommendations in parallel.
That is an interesting idea. @mhahsler what is your general view on using recommenderlab in production for data larger than movielense and other example data sets?
The code is not optimized for performance. Some recommendation algorithms will not scale well, others might...
Hello, thank you for such an awesome library. Can you help me to parallelize 'predict' function for get top N recommendation?