Wyniki wyszukiwania - Biblioteka Nauki

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

1 Fundamenta Informaticae

1 2014

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: lookahead in model tree induction

Sortuj według:

Ogranicz wyniki do:

Multi-Relational Model Tree Induction Tightly : Coupled with a Relational Database

100%

Appice A. , Ceci M. , Malerba D.

Fundamenta Informaticae

2014

tom Vol. 129, nr 3

193--224

Multi-Relational Data Mining (MRDM) refers to the process o f discovering implicit, previously unknown and potentially useful information fro m data scattered in multiple tables of a relational database. Following the mainstream of MRDM rese arch, we tackle the regression where the goal is to examine samples of past experience with known c ontinuous answers (response) and generalize future cases through an inductive process. Mr-S MOTI, the solution we propose, resorts to the structural approach in order to recursively partitio n data stored into a tightly-coupled database and build a multi-relational model tree which captures the l inear dependence between the response variable and one or more explanatory variables. The model tr ee is top-down induced by choosing, at each step, either to partition the training space or to intro duce a regression variable in the linear mod- els with the leaves. The tight-coupling with the database ma kes the knowledge on data structures (foreign keys) available free of charge to guide the search i n the multi-relational pattern space. Ex- periments on artificial and real databases demonstrate that in general Mr-SMOTI outperforms both SMOTI and M5’ which are two propositional model tree inducti on systems, and TILDE-RT which is a state-of-art structural model tree induction system