<!--
%\VignetteEngine{knitr::knitr}
%\VignetteIndexEntry{Non-personalized recommendations}
-->

## Non-personalized Recommendation in rrecsys

The included baseline recommendation methods are the global mean rating (Global Average), item’s mean rating (Item Average), user's mean rating (User Average), as well as an the Most Popular Item Recommendation method that recommends to all users the items that are most popular. 

<!--The "Most Mopular Item Recommendation" algorithm is the simplest implementation for item recommendations.
Every item will be ordered based on the number of times that it has been rated positively. 
A recommendation for a user will be the list of most popular items in the data set that the user has not chosen yet. -->

To train a model for the rating prediction task we use the following commands:

```{r, eval=FALSE}
data("ml100k")
d <- defineData(ml100k)
e <- evalModel(d, folds = 2)

globAv <- evalPred(e, alg = "globalaverage")
globAv

# Algorithm names might be matched on the registry partially.
itemAv <- evalPred(e,  "itemAver")
itemAv
userAv <- evalPred(e,  "useraverage")
userAv
```
Please notice that the baseline recommenders do not require any other argument.

The returned object are of type _algAverageClass_. 

To get more details about the slots read the reference [manual](https://CRAN.R-project.org/package=rrecsys).