This paper introduces a novel unsupervised text feature selection technique that combines the multi-objective evolutionary algorithm NSGA II with a local Hill Climbing based search. The objective functions in NSGA II are an adapted version of the Mean Absolute Difference criterion and the Holes projection pursuit index, both of which incorporate information regarding the cardinality of the selected feature subset. The local Hill Climbing procedure aims to further improve the solutions provided by NSGA II through elimination/replacement of redundant features. The redundancy level of a feature F in a chromosome is measured via the lexical and semantic similarities between F and the other features represented in the chromosome. Experiments conducted on three real-world textual subdatasets indicate that our proposed approach achieves very good results in comparison to four state-of-the-art evolutionary unsupervised feature selection techniques and to one classical feature extraction method with proven effectiveness (Principal Component Analysis).

Taking focused inspiration from biological evolution, we present an empirical study which shows that a Simple Genetic Algorithm (SGA) exhibits punctuated equilibria and punctuated gradualism in its evolution. Using the concept of consensus sequences, and comparing genotype change to phenotype change, we show how an SGA explores candidate solutions along a neutral network - Hamming-proximal bitstrings of similar fit-ness. Alongside mapping the normal functioning of an SGA, we monitor the formation of error thresholds “from above” by starting with a high mutation probability and slowly lowering it, during hundreds of thousands of generations. The formation of a stable consensus sequence is marked by a measurable upheaval in the dynamics of the population, leading to an efficient exploration of the search space in a short time. After the global optimum is found, we can still measure the degree of exploration the SGA performs on that neutral network, and observe punctuated equilibria. We use 11 numerical benchmark functions, along with the Royal Road Function, and a similar bit block Trap Function; the phenomena observed are largely similar on all of them, pointing to a generic behaviour of Genetic Algorithms, rather than problem particularities. Using a consensus sequence (a per-locus-mode chromosome) obscures quasispecies dynamics. This is why we use a per-locus-mean chromosome to measure information change between successive generations, and plot the number and maximal size of Quasispecies and Neutral Networks.

The widespread use of documents and email across the world, in various company departments (human resources, accounting, legal, etc) in conjunction with macro VBA language available for Office applications, has perpetuated a very efficient attack vector over the years. An attacker simply has to create a fake document with a macro VBA and send it to one of these departments masked as a legitimate document to engineer a potential breach. While security products have updated their detection technologies, so did the obfuscation techniques used in these attacks. This paper focuses on how this obfuscation techniques have evolved in the last decade.

Clustering is a fundamental and hence widely studied problem in data analysis. In a multi-objective perspective, this paper combines principles from two different clustering paradigms: the connectivity principle from density-based methods is integrated into the partitional clustering approach. The standard k-Means algorithm is hybridized with Particle Swarm Optimization. The new method (PSO-kMeans) benefits from both a local and a global view on data and alleviates some drawbacks of the k-Means algorithm; thus, it is able to spot types of clusters which are otherwise difficult to obtain (elongated shapes, non-similar volumes). Our experimental results show that PSO-kMeans improves the performance of standard k-Means in all test cases and performs at least comparable to state-of-the-art methods in the worst case. PSO-kMeans is robust to outliers. This comes at a cost: the preprocessing step for finding the nearest neighbors for each data item is required, which increases the initial linear complexity of k-Means to quadratic complexity.

Spin glasses are magnetic materials with strong disorder. Their unique properties assure them a central role in Statistical Physics, with applications that outreach many other important fields. Finding the ground state of an Ising spin glass is a highly multi-modal optimization problem, proved equivalent to the problem of finding the minimum weight cut in a graph. The states of a spin glass can be represented in terms of spins or bonds. Bond-based representations benefit from the symmetry properties of spin glasses, but require more memory due to the larger number of bonds. To tackle this issue, we introduced in our previous research a bond-based representation defined on spanning trees. A genetic algorithm (GA) using this representation (and special genetic operators) provided very good results for large problem instances (up to 125 spins). In this paper we continue this research. First, we describe how Particle Swarm Optimization (PSO) can be used to search ground states using the classical spin-based representation. Next, we present how the spanning tree representation can be used with PSO. Finally, we perform a comparative study of the performances of GAs and PSO with spin-based and bond-based representation on systems larger than before (up to 2500 spins).

Forecasting applications on the stock market attract much interest from researchers in the artificial intelligence field. The problem tackled in this study concerns predicting the direction of change of stock price indices, formulated in terms of binary classification. We use gene expression programming to evolve pools of binary classifiers and investigate several approaches to construct ensembles based on them. We compare the performance of the obtained classifiers with those of Naive Bayes, Support Vector Machines, Multilayer Perceptron, Decision Table and Random Forrest. The experiments performed on real-world stock market data show that the ensembles of GEP-evolved classifier models are competitive to classifiers trained by state-of-the-art machine learning methods.

The paper proposes a hypernetwork-based method for stock market prediction through a binary time series problem. Hypernetworks are a random hypergraph structure of higher-order probabilistic relations of data. The problem we tackle concerns the prediction of price movements (up/down) on stock markets. Compared to previous approaches, the proposed method discovers a large population of variable subpatterns, i.e. local and global patterns, using a novel evolutionary hypernetwork. An output is obtained from combining these patterns. In the paper, we describe two methods for assessing the prediction quality of the hypernetwork approach. Applied to the Dow Jones Industrial Average Index and the Korea Composite Stock Price Index data, the experimental results show that the proposed method effectively learns and predicts the time series information. In particular, the hypernetwork approach outperforms other machine learning methods such as support vector machines, naive Bayes, multilayer perceptrons, and k-nearest neighbors.

This paper is concerned with a problem in information organization and retrieval within Web communities. Most work in this domain is focused on reputation-based systems which exploit the experience gathered by previous users in order to evaluate resources at the community level. The current research focuses on a slightly different approach: a personalized evaluation system whose goal is to build a flexible and easy way to manage resources in a personalized manner. The functionality of such a model comes from local trust metrics which propagate the trust to a limited level into the system and, finally, lead to the appearance of minorities sharing some similar features/preferences. A modified PSO procedure is designed in order to analyze such a system and, in conjunction with a simple agglomerative clustering algorithm, identify homogenous groups of users.

This paper is concerned with feature weighting/selection in the context of unsupervised clustering. Since different subspaces of the feature space may lead to different partitions of the data set, an efficient algorithm to tackle multi-modal environments is needed. In this context, the Multi-Niche Crowding Genetic Algorithm is used for searching relevant feature subsets. The proposed method is designed to deal with the inherent biases regarding the number of clusters and the number of features that appear in an unsupervised framework. The first one is eliminated with the aid of a new unsupervised clustering criterion, while the second is tackled with the aid of cross-projection normalization. The method delivers a vector of weights which offers a ranking of features in accordance with their relevance to clustering.

This paper discusses ways of exploiting knowledge extracted from the optimization process, in order to assist an evolutionary solver in its path towards solution. In the context of NK fitness landscapes, we identify two facets of the difficulty of an optimization problem: the intrinsic combinatorial difficulty and its hybridization with random-search. For the experimental part, a particular case of NK fitness landscape is considered; traditional genetic algorithms and Integrated-Adaptive Genetic Algorithms (IAGA), which provide broad adaption mechanisms for most of GA’s components, are compared. We add to IAGA a learn-as-you-go system which allows operators to self-tune their behavior by inspecting the effect they produce on offspring. This system demonstrates that information derived from failures is as valuable as information obtained from positive experience. These results suggest that an appropriately designed adaptive system can be a tool for tackling problem difficulty caused by random-search hybridization.

This paper presents an attempt to transform PSO into a self-adaptive algorithm based on specific swarm-inspired operators. New features are introduced: spatial expansion intended to overcome premature convergence (an algorithm called improved PSO, IPSO) and auto-adaptation (an algorithm called adaptive PSO, APSO). Experiments show that APSO and IPSO outperform the basic PSO on benchmark problems, proving their efficiency especially on multimodal functions.

The aim of this paper is to present two heuristics vaguely inspired from the evolution of star systems. These methods are simple and can produce a good solution in a short time. While for small dimensional search spaces they can work alone, for large dimensional spaces their results can be used as an initial solution for some other heuristics. We study the effects of the initial solutions generated with these methods on the local search heuristics and on a genetic algorithm. Experimental results show that good solutions can be obtained with a combination of these methods.

Cross-clustering asks for a Boolean matrix to be brought to a quasi-canonical form. The problem has many applications in image processing, circuit design, archaeology, ecology etc. The heuristics currently used to solve it rely on either topological sorting or quasi-random search. We present here two evolutionary approaches to this problem: a permutation-based solution and a clustering one. The results on both real data and randomly generated, scalable, test data show very good convergence and encouraging efficiency properties, mainly for our second approach.

In a database where numeric data has been obtained by inaccurate methods (measurements, calculations involving error propagation) it is unlikely to have equal stored values for a single, repeatedly measured/calculated value. As a result, instances which should satisfy but actually "nearly" satisfy a functional dependency cannot be decomposed. For such cases, we introduce the notion of statistical dependency (sd) as an extension of functional dependencies (fds). We show how the well-known axioms for fds can be used in the case of sds; decompositions w.r.t. sds are also presented. The resulting decomposition can be used to answer statistical queries. Finally, we study a possible way of generalizing the multivalued dependencies in the same manner.

Answer-perturbation techniques for the protection of statistical databases have been previously introduced (Luchian and Stamate, 1992); they are flexible (perturbation kept under control), modular (do not interact with the DBMS) techniques, which compare favorably to previous protection techniques. In this paper, we generalise the answer-perturbation techniques w.r.t. the operation used for modifying the exact answers (thus enhancing the level of protection). Experimental results are also included; they indicate statistical soundness of our techniques.>