09.Machine learning09.Recommender systems - sporedata/researchdesigneR GitHub Wiki

1. Use cases: in which situations should I use this method?

It can be used for example, to make suggestions regarding treatments to cope with pain for cancer patients where, given a series of patient characteristics, similar patients have selected specific treatment options.

2. Input: what kind of data does the method require?
the same requisites required for the underlying machine learning model

3. Algorithm: how does the method work?

Model mechanics

The principle behind recommender systems such as non-negative matrix factorization is that it will check the relationship between a given patient and a matrix representing a combination of her characteristics and preferences, and then extrapolate those associations to other patients sharing the same characteristics but missing the preferences.

Reporting guidelines

Data science packages

Suggested companion methods

Learning materials

Books
Articles combining theory and scripts
- Learn how to implement Matrix Factorization algorithm that Google used in implementing collaborative filtering models [1].
- Common references for machine learning

4. Output: how do I interpret this method's results?

Mock conclusions or most frequent format for conclusions reached at the end of a typical analysis.

Tables, plots, and their interpretation

5. SporeData-specific

Templates

Data science functions

References

[1] Elshahawy S. Understanding Matrix Factorization for recommender systems.