Abstract
This paper presents a system for induction of forest of functional trees from data streams able to detect concept drift. The Ultra Fast Forest of Trees (UFFT) is an incremental algorithm, which works online, processing each example in constant time, and performing a single scan over the training examples. It uses analytical techniques to choose the splitting criteria, and the information gain to estimate the merit of each possible splitting-test. For multi-class problems the algorithm builds a binary tree for each possible pair of classes, leading to a forest of trees. Decision nodes and leaves contain naive-Bayes classifiers playing different roles during the induction process. Naive-Bayes in leaves are used to classify test examples. Naive-Bayes in inner nodes play two different roles. They can be used as multivariate splitting-tests if chosen by the splitting criteria, and used to detect changes in the class-distribution of the examples that traverse the node. When a change in the class-distribution is detected, all the sub-tree rooted at that node will be pruned. The use of naive-Bayes classifiers at leaves to classify test examples, the use of splitting-tests based on the outcome of naive-Bayes, and the use of naive-Bayes classifiers at decision nodes to detect changes in the distribution of the examples are directly obtained from the sufficient statistics required to compute the splitting criteria, without no additional computations. This aspect is a main advantage in the context of high-speed data streams. This methodology was tested with artificial and real-world data sets. The experimental results show a very good performance in comparison to a batch decision tree learner, and high capacity to detect drift in the distribution of the examples.

Abstract

Abstract
We address the problem of predicting the stability of secondary structure motifs of proteins given their linear sequence of residues. Our study is restricted to the prediction of helix structures. We have applied an Inductive Logic Programming (ILP) system to automatically synthesise the predictive rules. ILP systems are well known for being able to induce comprehensible models for data. Furthermore, the models components are definitions provided by a domain expert which makes the model more likely to be helpful in the understanding of the underlying process that produced the data. Our methodology has two stages. First, the system induces a model (set of rules) using just structural information and groupings of the residues to avoid biases by the domain expert. In the second stage, the residues properties are used to make the induced rules Chemically/Biologically appealing. We claim that this methodology is also valuable for general Structure-Activity Relationship (SAR) problems.

Abstract
Inductive Logic Programming (ILP) is an established subfield of Machine Learning. Nevertheless, it is recognized that efficiency and scalability is a major obstacle to an increased usage of ILP systems in complex applications with large hypotheses spaces. In this work, we focus on improving the efficiency and scalability of ILP systems by exploring tabling mechanisms available in the underlying Logic Programming systems. Tabling is an implementation technique that improves the declarativeness and performance of Prolog systems by reusing answers to subgoals. To validate our approach, we ran the April ILP system in the YapTab Prolog tabling system using two well-known datasets. The results obtained show quite impressive gains without changing the accuracy and quality of the theories generated.

Abstract
It is well known by Inductive Logic Programming (ILP) practioners that ILP systems usually take a long time to find valuable models (theories). The problem is specially critical for large datasets, preventing ILP systems to scale up to larger applications. One approach to reduce the execution time has been the parallelization of ILP systems. In this paper we overview the state-of-the-art on parallel ILP implementations and present work on the evaluation of some major parallelization strategies for ILP. Conclusions about the applicability of each strategy are presented.

Abstract
This paper aims at evaluate the impact of imitation networks on organizations' survival rates within a Portuguese industrial cluster. We used a Multi-Agent framework to represent the industrial cluster, its firms and the rules underlying the imitation strategies. Several experiments were based on the density dependence model, where vital rates are related with the size of the population (population density). We have concluded that imitation seems to improve the vital dynamics of the population and that present information about a firm is enough to establish an imitation network.