Feature subset selection for learning preferences

Abstract

In this paper we tackle a real world problem, the search of a function to evaluate the merits of beef cattle as meat producers. The independent variables represent a set of live animals' measurements; while the outputs cannot be captured with a single number, since the available experts tend to assess each animal in a relative way, comparing animals with the other partners in the same batch. Therefore, this problem can not be solved by means of regression methods; our approach is to learn the preferences of the experts when they order small groups of animals. Thus, the problem can be reduced to a binary classification, and can be dealt with a Support Vector Machine (SVM) improved with the use of a feature subset selection (FSS) method. We develop a method based on Recursive Feature Elimination (RFE) that employs an adaptation of a metric based method devised for model selection (ADJ). Finally, we discuss the extension of the resulting method to more general settings, and provide a comparison with other possible alternatives.

Analyzing Sensory Data Using Non-linear Preference Learning with Feature Subset Selection

“…In this paper we propose to tackle sensory data analysis by learning consumers' preferences, see [4,5,6] where training examples will be represented by preference judgments: pairs of vectors (v, u) where someone expresses that prefers the object represented by v to the object represented by u. We will show that this approach can induce more useful knowledge than other approaches, like regression based methods.…”

Dimension Reduction for Supervised Ordering

“…In [17,2], supervised ordering methods are used for sensory tests to examine which product features affect the value of the products. Metasearch engines are constructed in [4,8].…”

“…In the RCDR cases, the correlation vector size R (2) /N is much smaller than R (1) /N ; this means that the second vector is far less informative than the first, because the target ordering is generated by a linear function in this example. In the PCA case, the contribution ratio indicates that useful information still remains in this vector.…”

“…Some pre-processing techniques for mining or learning methods dealing with orders have been proposed. Bahamonde et al [2] applied wrapper-type feature selection to a supervised ordering task. Slotta et al [22] performed feature selection for classification of orders.…”

“…The first to fifth columns of each table show the components of vectors, w (1) and w (2) . In the last columns, values of …”