Ratings Class Reference

Evaluation class for rating prediction. More...

List of all members.

Static Public Member Functions
static double	ComputeCBD (double actual_rating, double prediction, double min_rating, double max_rating)
	Compute the capped binomial deviation (CBD)
static double	ComputeFit (this RatingPredictor recommender)
	Computes the RMSE fit of a recommender on the training data.
static RatingPredictionEvaluationResults	Evaluate (this IRatingPredictor recommender, IRatings test_ratings, IRatings training_ratings=null)
	Evaluates a rating predictor for RMSE, (N)MAE, and CBD.
Properties
static ICollection< string >	Measures [get]
	the evaluation measures for rating prediction offered by the class

---

Detailed Description

Evaluation class for rating prediction.

---

Member Function Documentation

static double ComputeCBD	(	double	actual_rating,
		double	prediction,
		double	min_rating,
		double	max_rating
	)		[inline, static]

Compute the capped binomial deviation (CBD)

http://www.kaggle.com/c/ChessRatings2/Details/Evaluation

Returns:

The CBD of a given rating and a prediction

Parameters:

actual_rating	the actual rating
prediction	the predicted rating
min_rating	the lower bound of the rating scale
max_rating	the upper bound of the rating scale

static double ComputeFit

(

this RatingPredictor

recommender

)

[inline, static]

Computes the RMSE fit of a recommender on the training data.

Returns:

the RMSE on the training data

Parameters:

recommender

the rating predictor to evaluate

static RatingPredictionEvaluationResults Evaluate	(	this IRatingPredictor	recommender,
		IRatings	test_ratings,
		IRatings	training_ratings = null
	)		[inline, static]

Evaluates a rating predictor for RMSE, (N)MAE, and CBD.

See http://recsyswiki.com/wiki/Root_mean_square_error and http://recsyswiki.com/wiki/Mean_absolute_error

For NMAE, see the paper by Goldberg et al.

For CBD (capped binomial deviance), see http://www.kaggle.com/c/ChessRatings2/Details/Evaluation

If the recommender can take time into account, and the rating dataset provides rating times, this information will be used for making rating predictions.

Literature:

• Ken Goldberg, Theresa Roeder, Dhruv Gupta, and Chris Perkins: Eigentaste: A Constant Time Collaborative Filtering Algorithm. nformation Retrieval Journal 2001. http://goldberg.berkeley.edu/pubs/eigentaste.pdf

Parameters:

recommender	rating predictor
test_ratings	test cases
training_ratings	the training examples

Returns:

a Dictionary containing the evaluation results

---

Property Documentation

ICollection<string> Measures [static, get]

the evaluation measures for rating prediction offered by the class

See http://recsyswiki.com/wiki/Root_mean_square_error and http://recsyswiki.com/wiki/Mean_absolute_error

---

The documentation for this class was generated from the following file:

• Eval/Ratings.cs