Stochastic decision predicates a scheme to represent motifs
Abstract: "This paper presents a new scheme for classifying genetic sequences, called Stochastic Decision Predicates. A stochastic decision predicate consists of Horn clauses and their probability parameters, and represents a (stochastic) motif that denotes a probabilistic mapping from a geneti...
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Buch |
| Sprache: | English |
| Veröffentlicht: |
Tokyo, Japan
1991
|
| Schriftenreihe: | Shin-Sedai-Konpyūta-Gijutsu-Kaihatsu-Kikō <Tōkyō>: ICOT technical report
657 |
| Schlagworte: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Zusammenfassung: | Abstract: "This paper presents a new scheme for classifying genetic sequences, called Stochastic Decision Predicates. A stochastic decision predicate consists of Horn clauses and their probability parameters, and represents a (stochastic) motif that denotes a probabilistic mapping from a genetic sequence to a set of categories, such as protein families. For the selection of stochastic decision predicates, quantative evaluation is possible from the viewpoint of predictive performance for unknown sequences as well as discrimination performance for the given genetic sequences. We employ Rissanen's Minimum Description Length (MDL) principle in order to avoid overlearning caused by the statistical fluctuation Our experimental results demonstrate that the MDL principle produces motifs with less predictive errors than the maximum likelihood method. |
|---|---|
| Beschreibung: | 7 S. |