Clustering algorithms for anti-money laundering using graph theory and social network analysis

HEMOLIA (a project under European community’s 7th framework programme) is a new generation Anti-Money Laundering (AML) intelligent multi-agent alert and investigation system which in addition to the traditional financial data makes extensive use of modern society’s huge telecom data source, thereby opening up a new dimension of capabilities to all Money Laundering fighters (FIUs, LEAs) and Financial Institutes (Banks, Insurance Companies, etc.). This Master-Thesis project is done at AIA, one of the partners for the HEMOLIA project in Barcelona. The objective of this thesis is to find the clusters in a network drawn by using the financial data. An extensive literature survey has been carried out and several standard algorithms related to networks have been studied and implemented. The clustering problem is a NP-hard problem and several algorithms like K-Means and Hierarchical clustering are being implemented for studying several problems relating to sociology, evolution, anthropology etc. However, these algorithms have certain drawbacks which make them very difficult to implement. The thesis suggests (a) a possible improvement to the K-Means algorithm, (b) a novel approach to the clustering problem using the Genetic Algorithms and (c) a new algorithm for finding the cluster of a node using the Genetic Algorithm.

Clustering compositional data trajectories

Our essay aims at studying suitable statistical methods for the clustering ofcompositional data in situations where observations are constituted by trajectories ofcompositional data, that is, by sequences of composition measurements along a domain.Observed trajectories are known as “functional data” and several methods have beenproposed for their analysis.In particular, methods for clustering functional data, known as Functional ClusterAnalysis (FCA), have been applied by practitioners and scientists in many fields. To ourknowledge, FCA techniques have not been extended to cope with the problem ofclustering compositional data trajectories. In order to extend FCA techniques to theanalysis of compositional data, FCA clustering techniques have to be adapted by using asuitable compositional algebra.The present work centres on the following question: given a sample of compositionaldata trajectories, how can we formulate a segmentation procedure giving homogeneousclasses? To address this problem we follow the steps described below.First of all we adapt the well-known spline smoothing techniques in order to cope withthe smoothing of compositional data trajectories. In fact, an observed curve can bethought of as the sum of a smooth part plus some noise due to measurement errors.Spline smoothing techniques are used to isolate the smooth part of the trajectory:clustering algorithms are then applied to these smooth curves.The second step consists in building suitable metrics for measuring the dissimilaritybetween trajectories: we propose a metric that accounts for difference in both shape andlevel, and a metric accounting for differences in shape only.A simulation study is performed in order to evaluate the proposed methodologies, usingboth hierarchical and partitional clustering algorithm. The quality of the obtained resultsis assessed by means of several indices

Solving Large Location-Allocation problems by Clustering and Simulated Annealing

Globalization involves several facility location problems that need to be handled at large scale. Location Allocation (LA) is a combinatorial problem in which the distance among points in the data space matter. Precisely, taking advantage of the distance property of the domain we exploit the capability of clustering techniques to partition the data space in order to convert an initial large LA problem into several simpler LA problems. Particularly, our motivation problem involves a huge geographical area that can be partitioned under overall conditions. We present different types of clustering techniques and then we perform a cluster analysis over our dataset in order to partition it. After that, we solve the LA problem applying simulated annealing algorithm to the clustered and non-clustered data in order to work out how profitable is the clustering and which of the presented methods is the most suitable

Discourse Type Clustering using POS n-gram Profiles and High-Dimensional Embeddings

Abstract: To cluster textual sequence types (discourse types/modes) in French texts, K-means algorithm with high-dimensional embeddings and fuzzy clustering algorithm were applied on clauses whose POS (part-ofspeech) n-gram profiles were previously extracted. Uni-, bi- and trigrams were used on four 19th century French short stories by Maupassant. For high-dimensional embeddings, power transformations on the chi-squared distances between clauses were explored. Preliminary results show that highdimensional embeddings improve the quality of clustering, contrasting the use of bi and trigrams whose performance is disappointing, possibly because of feature space sparsity.

BIO-INSPIRED COMPUTATIONAL TECHNIQUES APPLIED TO THE CLUSTERING AND VISUALIZATION OF SPATIO-TEMPORAL GEOSPATIAL DATA

The coverage and volume of geo-referenced datasets are extensive and incessantly¦growing. The systematic capture of geo-referenced information generates large volumes¦of spatio-temporal data to be analyzed. Clustering and visualization play a key¦role in the exploratory data analysis and the extraction of knowledge embedded in¦these data. However, new challenges in visualization and clustering are posed when¦dealing with the special characteristics of this data. For instance, its complex structures,¦large quantity of samples, variables involved in a temporal context, high dimensionality¦and large variability in cluster shapes.¦The central aim of my thesis is to propose new algorithms and methodologies for¦clustering and visualization, in order to assist the knowledge extraction from spatiotemporal¦geo-referenced data, thus improving making decision processes.¦I present two original algorithms, one for clustering: the Fuzzy Growing Hierarchical¦Self-Organizing Networks (FGHSON), and the second for exploratory visual data analysis:¦the Tree-structured Self-organizing Maps Component Planes. In addition, I present¦methodologies that combined with FGHSON and the Tree-structured SOM Component¦Planes allow the integration of space and time seamlessly and simultaneously in¦order to extract knowledge embedded in a temporal context.¦The originality of the FGHSON lies in its capability to reflect the underlying structure¦of a dataset in a hierarchical fuzzy way. A hierarchical fuzzy representation of¦clusters is crucial when data include complex structures with large variability of cluster¦shapes, variances, densities and number of clusters. The most important characteristics¦of the FGHSON include: (1) It does not require an a-priori setup of the number¦of clusters. (2) The algorithm executes several self-organizing processes in parallel.¦Hence, when dealing with large datasets the processes can be distributed reducing the¦computational cost. (3) Only three parameters are necessary to set up the algorithm.¦In the case of the Tree-structured SOM Component Planes, the novelty of this algorithm¦lies in its ability to create a structure that allows the visual exploratory data analysis¦of large high-dimensional datasets. This algorithm creates a hierarchical structure¦of Self-Organizing Map Component Planes, arranging similar variables' projections in¦the same branches of the tree. Hence, similarities on variables' behavior can be easily¦detected (e.g. local correlations, maximal and minimal values and outliers).¦Both FGHSON and the Tree-structured SOM Component Planes were applied in¦several agroecological problems proving to be very efficient in the exploratory analysis¦and clustering of spatio-temporal datasets.¦In this thesis I also tested three soft competitive learning algorithms. Two of them¦well-known non supervised soft competitive algorithms, namely the Self-Organizing¦Maps (SOMs) and the Growing Hierarchical Self-Organizing Maps (GHSOMs); and the¦third was our original contribution, the FGHSON. Although the algorithms presented¦here have been used in several areas, to my knowledge there is not any work applying¦and comparing the performance of those techniques when dealing with spatiotemporal¦geospatial data, as it is presented in this thesis.¦I propose original methodologies to explore spatio-temporal geo-referenced datasets¦through time. Our approach uses time windows to capture temporal similarities and¦variations by using the FGHSON clustering algorithm. The developed methodologies¦are used in two case studies. In the first, the objective was to find similar agroecozones¦through time and in the second one it was to find similar environmental patterns¦shifted in time.¦Several results presented in this thesis have led to new contributions to agroecological¦knowledge, for instance, in sugar cane, and blackberry production.¦Finally, in the framework of this thesis we developed several software tools: (1)¦a Matlab toolbox that implements the FGHSON algorithm, and (2) a program called¦BIS (Bio-inspired Identification of Similar agroecozones) an interactive graphical user¦interface tool which integrates the FGHSON algorithm with Google Earth in order to¦show zones with similar agroecological characteristics.

Tuning clustering in random networks with arbitrary degree distributions

We present a generator of random networks where both the degree-dependent clustering coefficient and the degree distribution are tunable. Following the same philosophy as in the configuration model, the degree distribution and the clustering coefficient for each class of nodes of degree k are fixed ad hoc and a priori. The algorithm generates corresponding topologies by applying first a closure of triangles and second the classical closure of remaining free stubs. The procedure unveils an universal relation among clustering and degree-degree correlations for all networks, where the level of assortativity establishes an upper limit to the level of clustering. Maximum assortativity ensures no restriction on the decay of the clustering coefficient whereas disassortativity sets a stronger constraint on its behavior. Correlation measures in real networks are seen to observe this structural bound.

Density-based hierarchical clustering of pyro-sequences on a large scale--the case of fungal ITS1.

MOTIVATION: Analysis of millions of pyro-sequences is currently playing a crucial role in the advance of environmental microbiology. Taxonomy-independent, i.e. unsupervised, clustering of these sequences is essential for the definition of Operational Taxonomic Units. For this application, reproducibility and robustness should be the most sought after qualities, but have thus far largely been overlooked. RESULTS: More than 1 million hyper-variable internal transcribed spacer 1 (ITS1) sequences of fungal origin have been analyzed. The ITS1 sequences were first properly extracted from 454 reads using generalized profiles. Then, otupipe, cd-hit-454, ESPRIT-Tree and DBC454, a new algorithm presented here, were used to analyze the sequences. A numerical assay was developed to measure the reproducibility and robustness of these algorithms. DBC454 was the most robust, closely followed by ESPRIT-Tree. DBC454 features density-based hierarchical clustering, which complements the other methods by providing insights into the structure of the data. AVAILABILITY: An executable is freely available for non-commercial users at ftp://ftp.vital-it.ch/tools/dbc454. It is designed to run under MPI on a cluster of 64-bit Linux machines running Red Hat 4.x, or on a multi-core OSX system. CONTACT: dbc454@vital-it.ch or nicolas.guex@isb-sib.ch.

Data-driven clustering : new methods and applications

Abstract : This work is concerned with the development and application of novel unsupervised learning methods, having in mind two target applications: the analysis of forensic case data and the classification of remote sensing images. First, a method based on a symbolic optimization of the inter-sample distance measure is proposed to improve the flexibility of spectral clustering algorithms, and applied to the problem of forensic case data. This distance is optimized using a loss function related to the preservation of neighborhood structure between the input space and the space of principal components, and solutions are found using genetic programming. Results are compared to a variety of state-of--the-art clustering algorithms. Subsequently, a new large-scale clustering method based on a joint optimization of feature extraction and classification is proposed and applied to various databases, including two hyperspectral remote sensing images. The algorithm makes uses of a functional model (e.g., a neural network) for clustering which is trained by stochastic gradient descent. Results indicate that such a technique can easily scale to huge databases, can avoid the so-called out-of-sample problem, and can compete with or even outperform existing clustering algorithms on both artificial data and real remote sensing images. This is verified on small databases as well as very large problems. Résumé : Ce travail de recherche porte sur le développement et l'application de méthodes d'apprentissage dites non supervisées. Les applications visées par ces méthodes sont l'analyse de données forensiques et la classification d'images hyperspectrales en télédétection. Dans un premier temps, une méthodologie de classification non supervisée fondée sur l'optimisation symbolique d'une mesure de distance inter-échantillons est proposée. Cette mesure est obtenue en optimisant une fonction de coût reliée à la préservation de la structure de voisinage d'un point entre l'espace des variables initiales et l'espace des composantes principales. Cette méthode est appliquée à l'analyse de données forensiques et comparée à un éventail de méthodes déjà existantes. En second lieu, une méthode fondée sur une optimisation conjointe des tâches de sélection de variables et de classification est implémentée dans un réseau de neurones et appliquée à diverses bases de données, dont deux images hyperspectrales. Le réseau de neurones est entraîné à l'aide d'un algorithme de gradient stochastique, ce qui rend cette technique applicable à des images de très haute résolution. Les résultats de l'application de cette dernière montrent que l'utilisation d'une telle technique permet de classifier de très grandes bases de données sans difficulté et donne des résultats avantageusement comparables aux méthodes existantes.

Aggregation invariance in general clustering approaches

General clustering deals with weighted objects and fuzzy memberships. We investigate the group- or object-aggregation-invariance properties possessed by the relevant functionals (effective number of groups or objects, centroids, dispersion, mutual object-group information, etc.). The classical squared Euclidean case can be generalized to non-Euclidean distances, as well as to non-linear transformations of the memberships, yielding the c-means clustering algorithm as well as two presumably new procedures, the convex and pairwise convex clustering. Cluster stability and aggregation-invariance of the optimal memberships associated to the various clustering schemes are examined as well.

Clustering of grape yield maps to delineate site-specific management zones

Zonal management in vineyards requires the prior delineation of stable yield zones within the parcel. Among the different methodologies used for zone delineation, cluster analysis of yield data from several years is one of the possibilities cited in scientific literature. However, there exist reasonable doubts concerning the cluster algorithm to be used and the number of zones that have to be delineated within a field. In this paper two different cluster algorithms have been compared (k-means and fuzzy c-means) using the grape yield data corresponding to three successive years (2002, 2003 and 2004), for a ‘Pinot Noir’ vineyard parcel. Final choice of the most recommendable algorithm has been linked to obtaining a stable pattern of spatial yield distribution and to allowing for the delineation of compact and average sized areas. The general recommendation is to use reclassified maps of two clusters or yield classes (low yield zone and high yield zone) and, consequently, the site-specific vineyard management should be based on the prior delineation of just two different zones or sub-parcels. The two tested algorithms are good options for this purpose. However, the fuzzy c-means algorithm allows for a better zoning of the parcel, forming more compact areas and with more equilibrated zonal differences over time.

Rockfall detection from LIDAR point clouds: A clustering approach using R

The analysis of rockfall characteristics and spatial distribution is fundamental to understand and model the main factors that predispose to failure. In our study we analysed LiDAR point clouds aiming to: (1) detect and characterise single rockfalls; (2) investigate their spatial distribution. To this end, different cluster algorithms were applied: 1a) Nearest Neighbour Clutter Removal (NNCR) in combination with the Expectation?Maximization (EM) in order to separate feature points from clutter; 1b) a density based algorithm (DBSCAN) was applied to isolate the single clusters (i.e. the rockfall events); 2) finally we computed the Ripley's K-function to investigate the global spatial pattern of the extracted rockfalls. The method allowed proper identification and characterization of more than 600 rockfalls occurred on a cliff located in Puigcercos (Catalonia, Spain) during a time span of six months. The spatial distribution of these events proved that rockfall were clustered distributed at a welldefined distance-range. Computations were carried out using R free software for statistical computing and graphics. The understanding of the spatial distribution of precursory rockfalls may shed light on the forecasting of future failures.

A new approach to segmentation based on fusing circumscribed contours, region growing and clustering

One of the major problems in machine vision is the segmentation of images of natural scenes. This paper presents a new proposal for the image segmentation problem which has been based on the integration of edge and region information. The main contours of the scene are detected and used to guide the posterior region growing process. The algorithm places a number of seeds at both sides of a contour allowing stating a set of concurrent growing processes. A previous analysis of the seeds permits to adjust the homogeneity criterion to the regions's characteristics. A new homogeneity criterion based on clustering analysis and convex hull construction is proposed

A concurrent region growing algorithm guided by circumscribed contours

Image segmentation of natural scenes constitutes a major problem in machine vision. This paper presents a new proposal for the image segmentation problem which has been based on the integration of edge and region information. This approach begins by detecting the main contours of the scene which are later used to guide a concurrent set of growing processes. A previous analysis of the seed pixels permits adjustment of the homogeneity criterion to the region's characteristics during the growing process. Since the high variability of regions representing outdoor scenes makes the classical homogeneity criteria useless, a new homogeneity criterion based on clustering analysis and convex hull construction is proposed. Experimental results have proven the reliability of the proposed approach

Management zones using fuzzy clustering based on spatial-temporal variability of soil and corn yield

Clustering soil and crop data can be used as a basis for the definition of management zones because the data are grouped into clusters based on the similar interaction of these variables. Therefore, the objective of this study was to identify management zones using fuzzy c-means clustering analysis based on the spatial and temporal variability of soil attributes and corn yield. The study site (18 by 250-m in size) was located in Jaboticabal, São Paulo/Brazil. Corn yield was measured in one hundred 4.5 by 10-m cells along four parallel transects (25 observations per transect) over five growing seasons between 2001 and 2010. Soil chemical and physical attributes were measured. SAS procedure MIXED was used to identify which variable(s) most influenced the spatial variability of corn yield over the five study years. Basis saturation (BS) was the variable that better related to corn yield, thus, semivariograms models were fitted for BS and corn yield and then, data values were krigged. Management Zone Analyst software was used to carry out the fuzzy c-means clustering algorithm. The optimum number of management zones can change over time, as well as the degree of agreement between the BS and corn yield management zone maps. Thus, it is very important take into account the temporal variability of crop yield and soil attributes to delineate management zones accurately.

Assessing the Reproducibility of Clustering of Molecular Dynamics Conformations on Self-Organizing Maps

The goal of most clustering algorithms is to find the optimal number of clusters (i.e. fewest number of clusters). However, analysis of molecular conformations of biological macromolecules obtained from computer simulations may benefit from a larger array of clusters. The Self-Organizing Map (SOM) clustering method has the advantage of generating large numbers of clusters, but often gives ambiguous results. In this work, SOMs have been shown to be reproducible when the same conformational dataset is independently clustered multiple times (~100), with the help of the Cramérs V-index (C_v). The ability of C_v to determine which SOMs are reproduced is generalizable across different SOM source codes. The conformational ensembles produced from MD (molecular dynamics) and REMD (replica exchange molecular dynamics) simulations of the penta peptide Met-enkephalin (MET) and the 34 amino acid protein human Parathyroid Hormone (hPTH) were used to evaluate SOM reproducibility. The training length for the SOM has a huge impact on the reproducibility. Analysis of MET conformational data definitively determined that toroidal SOMs cluster data better than bordered maps due to the fact that toroidal maps do not have an edge effect. For the source code from MATLAB, it was determined that the learning rate function should be LINEAR with an initial learning rate factor of 0.05 and the SOM should be trained by a sequential algorithm. The trained SOMs can be used as a supervised classification for another dataset. The toroidal 10×10 hexagonal SOMs produced from the MATLAB program for hPTH conformational data produced three sets of reproducible clusters (27%, 15%, and 13% of 100 independent runs) which find similar partitionings to those of smaller 6×6 SOMs. The χ^2 values produced as part of the C_v calculation were used to locate clusters with identical conformational memberships on independently trained SOMs, even those with different dimensions. The χ^2 values could relate the different SOM partitionings to each other.