Itamar Hata

The Occam’s Razor principle has become the basis for many Machine Learning algorithms, under the interpretation that the classifier should not be more complex than necessary. Recently, this principle has shown to be well suited to associative classifiers, where the number of rules composing the classifier can be substantially reduced by using condensed representations such as maximal or closed rules. While it is shown that such a decrease in the complexity of the classifier (usually) does not… Mostrar más

The Occam’s Razor principle has become the basis for many Machine Learning algorithms, under the interpretation that the classifier should not be more complex than necessary. Recently, this principle has shown to be well suited to associative classifiers, where the number of rules composing the classifier can be substantially reduced by using condensed representations such as maximal or closed rules. While it is shown that such a decrease in the complexity of the classifier (usually) does not compromise its accuracy, the number of remaining rules is still larger than necessary and making it hard for experts to interpret the corresponding classifier. In this paper we propose a much more aggressive filtering strategy, which decreases the number of rules within the classifier dramatically without hurting its accuracy. Our strategy consists in evaluating each rule under different statistical criteria, and filtering only those rules that show a positive balance between all the criteria considered. Specifically, each candidate rule is associated with a
point in an n-dimensional scattergram, where each coordinate corresponds to a statistical criterion. Points that are not dominated by any other point in the scattergram compose the Pareto frontier, and correspond to rules that are optimal in the sense that there is no rule that is better off when all the criteria are taken into . Finally, rules lying in the Pareto frontier are filtered and compose the classifier. A systematic set of experiments involving benchmark data as well as recent data from actual application scenarios, followed by an extensive set of significance tests, reveal that the proposed strategy decreases the number of rules by up to two orders of magnitude and produces classifiers that are extremely readable (i.e., allow interpretability of the classification results) without hurting accuracy.
Mostrar menos

In this paper we propose new approaches for multi-objective recommender systems based on the concept of Pareto-efficiency − a state achieved when the system is devised in the
most efficient manner in the sense that there is no way to improve one of the objectives without making
any other objective worse off. Given that existing multi-objective recommendation algorithms differ in their
level of accuracy, diversity and novelty, we exploit the Pareto-efficiency concept in two distinct… Mostrar más

In this paper we propose new approaches for multi-objective recommender systems based on the concept of Pareto-efficiency − a state achieved when the system is devised in the
most efficient manner in the sense that there is no way to improve one of the objectives without making
any other objective worse off. Given that existing multi-objective recommendation algorithms differ in their
level of accuracy, diversity and novelty, we exploit the Pareto-efficiency concept in two distinct manners:
(i) the aggregation of ranked lists produced by existing algorithms into a single one, which we call Pareto-efficient ranking, and (ii) the weighted combination of existing algorithms resulting in a hybrid one, which
we call Pareto-efficient hybridization. Our evaluation involves two real application scenarios: music recommendation with implicit (i.e., Last.fm) and movie recommendation with explicit (i.e., MovieLens). We show that the proposed Pareto-efficient approaches are effective in suggesting items that
are likely to be simultaneously accurate, diverse and novel. We discuss scenarios where the system achieves
high levels of diversity and novelty without compromising its accuracy. Further, comparison against multi-
objective baselines reveals improvements in of accuracy (from 10.4% to 10.9%), novelty (from 5.7% to 7.5%), and diversity(from 1.6% to 4.2%).
Mostrar menos

Performance modeling is a central task in the capacity management of server platforms. Traditional performance models are created for transactional workloads (purely open models), batch or interactive (purely closed models). However, many real Web applications experience session-based workloads that exhibit a partially-open behavior, which includes components from either open or closed model, particularly with respect to response time. In this paper, we developed an analytical model for… Mostrar más

Performance modeling is a central task in the capacity management of server platforms. Traditional performance models are created for transactional workloads (purely open models), batch or interactive (purely closed models). However, many real Web applications experience session-based workloads that exhibit a partially-open behavior, which includes components from either open or closed model, particularly with respect to response time. In this paper, we developed an analytical model for performance that captures the main aspects
of partially-open workloads of multi-layered Web platforms and compared it with the open and closed models by simulation. Mostrar menos

Gerenciamento de capacidade de uma infra-estrutura de hospedagem
tem tradicionalmente focado em questões de desempenho. Entretanto, a qualidade
do serviçoo oferecido às aplicações hospedadas e o lucro do provedor dependem
de outros aspectos, como segurança e restrições de energia. Este artigo
estende nossa solução de gerenciamento de capacidade auto-adaptativo para
capturar os compromissos chave de desempenho e custos que surgem quando
aplicações são alvos de ataques de… Mostrar más

Gerenciamento de capacidade de uma infra-estrutura de hospedagem
tem tradicionalmente focado em questões de desempenho. Entretanto, a qualidade
do serviçoo oferecido às aplicações hospedadas e o lucro do provedor dependem
de outros aspectos, como segurança e restrições de energia. Este artigo
estende nossa solução de gerenciamento de capacidade auto-adaptativo para
capturar os compromissos chave de desempenho e custos que surgem quando
aplicações são alvos de ataques de segurança e a infra-estrutura sofre restrições
de energia. Vários cenários e estratégias baseadas em SLAs dinâmicos foram
formuladas para descobrir os principais compromissos de custos e desempenho,
considerando tanto os interesses do provedor quanto do cliente. Mostrar menos

Capacity management of a hosting infrastructure
has traditionally focused only on performance goals. However,
the quality of service provided to the hosted applications, and
ultimately the revenues achieved by the provider, depend also on
other aspects, such as security and energy constraints. This paper
extends our self-adaptive SLA-driven capacity management solution
to capture, in an unified framework, key performance and
cost tradeoffs that arise when operating under… Mostrar más

Capacity management of a hosting infrastructure
has traditionally focused only on performance goals. However,
the quality of service provided to the hosted applications, and
ultimately the revenues achieved by the provider, depend also on
other aspects, such as security and energy constraints. This paper
extends our self-adaptive SLA-driven capacity management solution
to capture, in an unified framework, key performance and
cost tradeoffs that arise when operating under security attacks
and energy constraints. A number of scenarios and strategies
based on dynamic SLA contracts are designed to help uncover, via
simulation experiments, the main tradeoffs, considering both the
provider’s interests (i.e., revenues) and the customer’s interests
(i.e., legitimate throughput, response time distribution and costs).
Finally, we also assess the cost-effectiveness of our framework
under highly variable application service times. Mostrar menos