AI*IA 2001: Advances in Artificial Intelligence: 7th by A. Serra, A. Giordana (auth.), Floriana Esposito (eds.)

By A. Serra, A. Giordana (auth.), Floriana Esposito (eds.)

This e-book constitutes the refereed court cases of the clinical song of the seventh Congress of the Italian organization for man made Intelligence, AI*IA 2001, held in Bari, Italy, in September 2001.
The 25 revised lengthy papers and sixteen revised brief papers have been conscientiously reviewed and chosen for inclusion within the quantity. The papers are prepared in topical sections on laptop studying; computerized reasoning; wisdom illustration; multi-agent structures; average language processing; conception, imaginative and prescient, and robotics; and making plans and scheduling.

Show description

Read Online or Download AI*IA 2001: Advances in Artificial Intelligence: 7th Congress of the Italian Association for Artificial Intelligence Bari, Italy, September 25–28, 2001 Proceedings PDF

Similar italian books

Italiano in Cinque Minuti: Esercizi Rapidi Sulla Grammatica e Sul Lessico Volume 1

Грамматические и лексические упражнения по итальянскому языку. 1. Nomi e pronomi. 2. Articoli. three. Presente dei verbi regolari. four. Essere, avere, esserci. five. Aggettivi qualificativi. 6. Forma di cortesia. 7. Aggettivi e pronomi possessivi. eight. Presente dei verbi irregolari e modali. nine. Aggettivi e pronomi dimostrativi.

Additional resources for AI*IA 2001: Advances in Artificial Intelligence: 7th Congress of the Italian Association for Artificial Intelligence Bari, Italy, September 25–28, 2001 Proceedings

Example text

It assigns to all hyperlinks inside a document the same relevance, given by the class a posteriori probability of the document containing them. We call this method neighbourhood score, because it is based on the assumption that relevant pages for a given class will point to other relevant pages for the same class. We calculate such a probability by the extended Naive Bayes classifier described in previous section. The second method, called anchor score, assigns each hyperlink a different score depending on the text of its anchor, that is the text that can be clicked when viewing the document with a browser, or the text contained in the ALT tag in case of a clickable image.

Some of the model tree induction systems developed are: M5 [8], RETIS [4], M5' [12], RegTree [5], and HTL [10,11]. All these systems perform a top-down induction of models trees (TDIMT) by building the tree structure through recursive partitioning of the training set and by associating leaves with models. During the construction of the tree there are three main problems to be solved: Choosing the best partition of a region of the feature space, determining the leaves of the tree and choosing a model for each leaf.

The network topology is comprised of three layers with the following meaning: 1. The first layer calculates the membership values µik (xi ). Each node of this layer has 2 free parameters (center and width of Gaussian function). 2. The second layer computes the activation strength of each rule, by multiplying the corresponding input membership values. Nodes of this layer have no free parameters. 42 Giovanna Castellano and Anna Maria Fanelli 3. The third layer determines the output values yj , j = 1 .

Download PDF sample

Rated 4.12 of 5 – based on 44 votes