Associative classifcation is a promising technique for the generalion of highly precise classifers. Prpvious works propose several clever rechniques ro prune rhe huge ser of generared rules, with rhe rwofold aim of selecring a small ser of high quality rules. and reducing the chance of eve@ I n rhis papez we argue rhar pruning should be reduced ro a minimum and rhar rhe ovailability of a large rule base may improve the precision of rke classifer: wirhour affecting its performance. In L:' (Live and Let Live), a new algorirhm for associarive classifcarion, o lazy pruning technique irerarively discards all rules rhar only yield wrong case classificarions. Classifcarion is performed in rwo steps. Inirially, rules which have already correcrly classified at leasr one training case, sorred by confidence, are considered. Ifthe case is srill unclassijed, rhe remaining rules (unused during rhe rroining phasej are considered, again sorred by confidence. Exrensive experimenrs on 26 darabases fmm rhe U C I machine learning darabase reposirory show thar L7 improves rhe classifcarion precision with respecr ro previous approaches. ring.
Rules in active database systems can be very difficult to program due to the unstructured and unpredictable nature of rule processing. We provide static analysis techniques for predicting whether a given rule set is guaranteed to terminate and whether rule execution is confluent (guaranteed to have a unique final state). Our methods are based on previous techniques for analyzing rules in active database systems. We improve considerably on the previous techniques by providing analysis criteria that are much less conservative: our methods often determine that a rule set will terminate or is confluent when previous methods could not make this determination. Our improved analysis is based on a "propagation" algorithm, which uses an extended relational algebra to accurately determine when the action of one rule can affect the condition of another, and determine when rule actions commute. We consider both condition-action rules and event-condition-action rules, making our approach widely applicable to relational active database rule languages and to the trigger language in the SQL:1999 standard.
Association rule extraction is a widely used exploratory technique which has been exploited in different contexts (e.g., biological data, medical images). However, association rule extraction, driven by support and confidence constraints, entails (i) generating a huge number of rules which are difficult to analyze, or (ii) pruning rare itemsets, even if their hidden knowledge might be relevant. To address the above issues, this paper presents a novel frequent itemset mining algorithm, called GENIO (GENeralized Itemset DiscOverer), to analyze correlation among data by means of generalized itemsets, which provide a powerful tool to efficiently extract hidden knowledge, discarded by previous approaches. The proposed technique exploits a (user provided) taxonomy to drive the pruning phase of the extraction process. Instead of extracting itemsets for all levels of the taxonomy and post-pruning them, the GenIO algorithm performs a support driven opportunistic aggregation of itemsets. Generalized itemsets are extracted only if itemsets at a lower level in the taxonomy are below the support threshold. Experiments performed in the network traffic domain show the efficiency and the effectiveness of the proposed algorithm
This paper presents a flexible framework that performs real-time analysis of physiological data to monitor people's health conditions in any context (e.g., during daily activities, in hospital environments). Given historical physiological data, different behavioral models tailored to specific conditions (e.g., a particular disease, a specific patient) are automatically learnt. A suitable model for the currently monitored patient is exploited in the real-time stream classification phase. The framework has been designed to perform both instantaneous evaluation and stream analysis over a sliding time window. To allow ubiquitous monitoring, real-time analysis could also be executed on mobile devices. As a case study, the framework has been validated in the intensive care scenario. Experimental validation, performed on 64 patients affected by different critical illnesses, demonstrates the effectiveness and the flexibility of the proposed framework in detecting different severity levels of monitored people's clinical situations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.