Recommender System - AbhayAg/MachineLearning GitHub Wiki
- For saving memory we should use broadcast variables if some variables are getting used multiple times.
- In case of algorithm if one variable is independent of other then we should use parallelization to save time.