Results 1  10
of
52
The Similarity Metric
 IEEE TRANSACTIONS ON INFORMATION THEORY
, 2003
"... A new class of distances appropriate for measuring similarity relations between sequences, say one type of similarity per distance, is studied. We propose a new "normalized information distance", based on the noncomputable notion of Kolmogorov complexity, and show that it is in this class and it min ..."
Abstract

Cited by 186 (26 self)
 Add to MetaCart
A new class of distances appropriate for measuring similarity relations between sequences, say one type of similarity per distance, is studied. We propose a new "normalized information distance", based on the noncomputable notion of Kolmogorov complexity, and show that it is in this class and it minorizes every computable distance in the class (that is, it is universal in that it discovers all computable similarities). We demonstrate that it is a metric and call it the similarity metric. This theory forms the foundation for a new practical tool. To evidence generality and robustness we give two distinctive applications in widely divergent areas using standard compression programs like gzip and GenCompress. First, we compare whole mitochondrial genomes and infer their evolutionary history. This results in a first completely automatic computed whole mitochondrial phylogeny tree. Secondly, we fully automatically compute the language tree of 52 different languages.
Towards parameterfree data mining
 In: Proc. 10th ACM SIGKDD Intn’l Conf. Knowledge Discovery and Data Mining
, 2004
"... Most data mining algorithms require the setting of many input parameters. Two main dangers of working with parameterladen algorithms are the following. First, incorrect settings may cause an algorithm to fail in finding the true patterns. Second, a perhaps more insidious problem is that the algorit ..."
Abstract

Cited by 118 (18 self)
 Add to MetaCart
Most data mining algorithms require the setting of many input parameters. Two main dangers of working with parameterladen algorithms are the following. First, incorrect settings may cause an algorithm to fail in finding the true patterns. Second, a perhaps more insidious problem is that the algorithm may report spurious patterns that do not really exist, or greatly overestimate the significance of the reported patterns. This is especially likely when the user fails to understand the role of parameters in the data mining process. Data mining algorithms should have as few parameters as possible, ideally none. A parameterfree algorithm would limit our ability to impose our prejudices, expectations, and presumptions on the problem at hand, and would let the data itself speak to us. In this work, we show that recent results in bioinformatics and computational theory hold great promise for a parameterfree datamining paradigm. The results are motivated by observations in Kolmogorov complexity theory. However, as a practical matter, they can be implemented using any offtheshelf compression algorithm with the addition of just a dozen or so lines of code. We will show that this approach is competitive or superior to the stateoftheart approaches in anomaly/interestingness detection, classification, and clustering with empirical tests on time series/DNA/text/video datasets.
Shared Information and Program Plagiarism Detection
 IEEE TRANS. INFORM. TH
"... A fundamental question in information theory and in computer science is how to measure similarity or the amount of shared information between two sequences. We have proposed a metric, based on Kolmogorov complexity to answer this question, and have proven it to be universal. We apply this metric i ..."
Abstract

Cited by 62 (2 self)
 Add to MetaCart
A fundamental question in information theory and in computer science is how to measure similarity or the amount of shared information between two sequences. We have proposed a metric, based on Kolmogorov complexity to answer this question, and have proven it to be universal. We apply this metric in measuring the amount of shared information between two computer programs, to enable plagiarism detection. We have
Alignmentfree sequence comparisona review
 Bioinformatics
, 2003
"... Motivation: Genetic recombination and, in particular, genetic shuffling are at odds with sequence comparison by alignment, which assumes conservation of contiguity between homologous segments. A variety of theoretical foundations are being used to derive alignmentfree methods that overcome this lim ..."
Abstract

Cited by 42 (5 self)
 Add to MetaCart
Motivation: Genetic recombination and, in particular, genetic shuffling are at odds with sequence comparison by alignment, which assumes conservation of contiguity between homologous segments. A variety of theoretical foundations are being used to derive alignmentfree methods that overcome this limitation. The formulation of alternative metrics for dissimilarity between sequences and their algorithmic implementations are reviewed. Results: The overwhelming majority of work on alignmentfree sequence has taken place in the past two decades, with most reports published in the past 5 years. Two main categories of methods have been proposed—methods based on word (oligomer) frequency, and methods that do not require resolving the sequence with fixed word length segments. The first category is based on the statistics of word frequency, on the distances defined in a Cartesian space defined by the frequency vectors, and on the information content of frequency distribution. The second category includes the use of Kolmogorov complexity and Chaos Theory. Despite their low visibility, alignmentfree metrics are in fact already widely used as preselection filters for alignmentbased querying of large applications. Recent work is furthering their usage as a scaleindependent methodology that is capable of recognizing homology when loss of contiguity is beyond the possibility of alignment. Availability: Most of the alignmentfree algorithms reviewed were implemented in MATLAB code and are available
Inferring Phylogenetic Trees Using Evolutionary Algorithms
 Parallel Problem Solving From Nature VII
, 2002
"... We consider the problem of estimating the evolutionary history of a collection of organisms in terms of a phylogenetic tree. This is a hard combinatorial optimization problem for which different EA approaches are proposed and evaluated. Using two problem instances of different sizes, it is shown tha ..."
Abstract

Cited by 15 (6 self)
 Add to MetaCart
We consider the problem of estimating the evolutionary history of a collection of organisms in terms of a phylogenetic tree. This is a hard combinatorial optimization problem for which different EA approaches are proposed and evaluated. Using two problem instances of different sizes, it is shown that an EA that directly encodes trees and uses adhoc operators performs better than several decoderbased EAs, but does not scale well with the problem size. A greedydecoder EA provides the overall best results, achieving near 100%success at a lower computational cost than the remaining approaches.
Algorithmic Complexity
 M B
, 1993
"... The theory of algorithmic complexity (commonly known as Kolmogorov complexity) or algorithmic information theory is a novel mathematical approach combining the theory of computation with information theory. It is the theory that finally formalizes the elusive notion of the amount of information in i ..."
Abstract

Cited by 14 (9 self)
 Add to MetaCart
The theory of algorithmic complexity (commonly known as Kolmogorov complexity) or algorithmic information theory is a novel mathematical approach combining the theory of computation with information theory. It is the theory that finally formalizes the elusive notion of the amount of information in individual objects, in contrast to entropy that is a statistical notion of average code word length to transmit a message form a random source. This powerful new theory has successfully resolved ancient questions about the nature of randomness of individual objects, inductive reasoning and prediction, and has applications in mathematics, computer science, physics, biology, and other sciences, including social and behavioral sciences.
Causal inference using the algorithmic Markov condition
, 2008
"... Inferring the causal structure that links n observables is usually basedupon detecting statistical dependences and choosing simple graphs that make the joint measure Markovian. Here we argue why causal inference is also possible when only single observations are present. We develop a theory how to g ..."
Abstract

Cited by 11 (11 self)
 Add to MetaCart
Inferring the causal structure that links n observables is usually basedupon detecting statistical dependences and choosing simple graphs that make the joint measure Markovian. Here we argue why causal inference is also possible when only single observations are present. We develop a theory how to generate causal graphs explaining similarities between single objects. To this end, we replace the notion of conditional stochastic independence in the causal Markov condition with the vanishing of conditional algorithmic mutual information anddescribe the corresponding causal inference rules. We explain why a consistent reformulation of causal inference in terms of algorithmic complexity implies a new inference principle that takes into account also the complexity of conditional probability densities, making it possible to select among Markov equivalent causal graphs. This insight provides a theoretical foundation of a heuristic principle proposed in earlier work. We also discuss how to replace Kolmogorov complexity with decidable complexity criteria. This can be seen as an algorithmic analog of replacing the empirically undecidable question of statistical independence with practical independence tests that are based on implicit or explicit assumptions on the underlying distribution. email:
Compression and machine learning: A new perspective on feature space vectors
 in To Appear in Data Compression Conference (DCC
, 2006
"... The use of compression algorithms in machine learning tasks such as clustering and classification has appeared in a variety of fields, sometimes with the promise of reducing problems of explicit feature selection. The theoretical justification for such methods has been founded on an upper bound on K ..."
Abstract

Cited by 11 (2 self)
 Add to MetaCart
The use of compression algorithms in machine learning tasks such as clustering and classification has appeared in a variety of fields, sometimes with the promise of reducing problems of explicit feature selection. The theoretical justification for such methods has been founded on an upper bound on Kolmogorov complexity and an idealized information space. An alternate view shows compression algorithms implicitly map strings into implicit feature space vectors, and compressionbased similarity measures compute similarity within these feature spaces. Thus, compressionbased methods are not a “parameter free ” magic bullet for feature selection and data representation, but are instead concrete similarity measures within defined feature spaces, and are therefore akin to explicit feature vector models used in standard machine learning algorithms. To underscore this point, we find theoretical and empirical connections between traditional machine learning vector models and compression, encouraging crossfertilization in future work. 1. Introduction: ReInventing the Razor The fundamental idea that data compression can be used to perform machine learning tasks has surfaced in a several areas of research, including data compression (Witten et al., 1999a; Frank et al., 2000), machine learning and data mining (Cilibrasi and Vitanyi, 2005; Keogh et al., 2004;
A Simple Statistical Algorithm for Biological Sequence Compression
 DATA COMPRESSION CONFERENCE
, 2007
"... This paper introduces a novel algorithm for biological sequence compression that makes use of both statistical properties and repetition within sequences. A panel of experts is maintained to estimate the probability distribution of the next symbol in the sequence to be encoded. Expert probabilities ..."
Abstract

Cited by 11 (0 self)
 Add to MetaCart
This paper introduces a novel algorithm for biological sequence compression that makes use of both statistical properties and repetition within sequences. A panel of experts is maintained to estimate the probability distribution of the next symbol in the sequence to be encoded. Expert probabilities are combined to obtain the final distribution. The resulting information sequence provides insight for further study of the biological sequence. Each symbol is then encoded by arithmetic coding. Experiments show that our algorithm outperforms existing compressors on typical DNA and protein sequence datasets while maintaining a practical running time. 1.
Compressed qgram Indexing for Highly Repetitive Biological Sequences
"... Abstract—The study of compressed storage schemes for highly repetitive sequence collections has been recently boosted by the availability of cheaper sequencing technologies and the flood of data they promise to generate. Such a storage scheme may range from the simple goal of retrieving whole indivi ..."
Abstract

Cited by 10 (8 self)
 Add to MetaCart
Abstract—The study of compressed storage schemes for highly repetitive sequence collections has been recently boosted by the availability of cheaper sequencing technologies and the flood of data they promise to generate. Such a storage scheme may range from the simple goal of retrieving whole individual sequences to the more advanced one of providing fast searches in the collection. In this paper we study alternatives to implement a particularly popular index, namely, the one able of finding all the positions in the collection of substrings of fixed length (qgrams). We introduce two novel techniques and show they constitute practical alternatives to handle this scenario. They excell particularly in two cases: when q is small (up to 6), and when the collection is extremely repetitive (less than 0.01 % mutations). I. INTRODUCTION AND RELATED WORK The sequencing of the whole Human Genome was a celebrated