Luisa Micó

A fast branch & bound nearest neighbour classifier in metric spaces

The recently introduced algorithm LAESA finds the nearest neighbour prototype in a metric space. The average number of distances computed in the algorithm does not depend on the number of prototypes but it shows linear space and time complexities. In this paper, a new algorithm (TLAESA) is proposed which has a sublinear time complexity and keeps the other features unchanged.

Comparison of classifier fusion methods for classification in pattern recognition tasks

This work presents a comparison of current research in the use of voting ensembles of classifiers in order to improve the accuracy of single classifiers and make the performance more robust against the difficulties that each individual classifier may have. Also, a number of combination rules are proposed. Different voting schemes are discussed and compared in order to study the performance of the ensemble in each task. The ensembles have been trained on real data available for benchmarking and also applied to a case study related to statistical description models of melodies for music genre recognition.

Comparison of AESA and LAESA search algorithms using string and tree-edit-distances

Although the success rate of handwritten character recognition using a nearest neighbour technique together with edit distance is satisfactory, the exhaustive search is expensive. Some fast methods as AESA and LAESA have been proposed to find nearest neighbours in metric spaces. The average number of distances computed by these algorithms is very low and does not depend on the number of prototypes in the training set. In this paper, we compare the behaviour of these algorithms when string and tree-edit-distances are used.

Some approaches to improve tree-based nearest neighbour search algorithms

Nearest neighbour search is a widely used technique in pattern recognition. During the last three decades a large number of fast algorithms have been proposed. In this work we are interested in algorithms that can be used with any dissimilarity function provided that it fits the mathematical notion of distance. Some of such algorithms organize, in preprocessing time, the data in a tree structure that is traversed in search time to find the nearest neighbour. The speedup is obtained using some pruning rules that avoid the traversal of some parts of the tree. In this work two new decomposition methods to build the tree and three new pruning rules are explored. The behaviour of our proposal is studied through experiments with synthetic and real data.

Comparison of fast nearest neighbour classifiers for handwritten character recognition

Abstract Recently some fast methods ( LAESA and TLAESA ) have been proposed to find nearest neighbours in metric spaces. The average number of distances computed by these algorithms does not depend on the number of prototypes and they show linear space complexity. These results where obtained through vast experimentation using only artificial data. In this paper, we corroborate this behaviour when applied to handwritten character recognition tasks. Moreover, we compare LAESA and TLAESA with some classical algorithms also working in metric spaces.

An approximate median search algorithm in non-metric spaces

Abstract Given a set of data points and a distance function, the median point is defined as the point (in the set) that minimizes the sum of the distances to the remaining points of the set. In the general case, the median computation has an O( n 2 ) time cost, where n is the number of points. Nevertheless, for most tasks an approximate median is enough. In this paper a very fast algorithm (linear in time) is presented that finds a point that is a very good approximation of the exact median. This algorithm is independent of the distance function and does not degrade as the dimensionality of the data increases.

Extending LAESA Fast Nearest Neighbour Algorithm to Find the k Nearest Neighbours

Many pattern recognition tasks make use of the k nearest neighbour (k-NN) technique. In this paper we are interested on fast k- NN search algorithms that can work in any metric space i.e. they are not restricted to Euclidean-like distance functions. Only symmetric and triangle inequality properties are required for the distance.A large set of such fast k-NN search algorithms have been developed during last years for the special case where k = 1. Some of them have been extended for the general case. This paper proposes an extension of LAESA (Linear Approximation Elimination Search Algorithm) to find the k-NN.

Music staff removal with supervised pixel classification

This work presents a novel approach to tackle the music staff removal. This task is devoted to removing the staff lines from an image of a music score while maintaining the symbol information. It represents a key step in the performance of most optical music recognition systems. In the literature, staff removal is usually solved by means of image processing procedures based on the intrinsics of music scores. However, we propose to model the problem as a supervised learning classification task. Surprisingly, although there is a strong background and a vast amount of research concerning machine learning, the classification approach has remained unexplored for this purpose. In this context, each foreground pixel is labelled as either staff or symbol. We use pairs of scores with and without staff lines to train classification algorithms. We test our proposal with several well-known classification techniques. Moreover, in our experiments no attempt of tuning the classification algorithms has been made, but the parameters were set to the default setting provided by the classification software libraries. The aim of this choice is to show that, even with this straightforward procedure, results are competitive with state-of-the-art algorithms. In addition, we also discuss several advantages of this approach for which conventional methods are not applicable such as its high adaptability to any type of music score.

A fast pivot-based indexing algorithm for metric spaces

This work focus on fast nearest neighbor (NN) search algorithms that can work in any metric space (not just the Euclidean distance) and where the distance computation is very time consuming. One of the most well known methods in this field is the AESA algorithm, used as baseline for performance measurement for over twenty years. The AESA works in two steps that repeats: first it searches a promising candidate to NN and computes its distance (approximation step), next it eliminates all the unsuitable NN candidates in view of the new information acquired in the previous calculation (elimination step). This work introduces the PiAESA algorithm. This algorithm improves the performance of the AESA algorithm by splitting the approximation criterion: on the first iterations, when there is not enough information to find good NN candidates, it uses a list of pivots (objects in the database) to obtain a cheap approximation of the distance function. Once a good approximation is obtained it switches to the AESA usual behavior. As the pivot list is built in preprocessing time, the run time of PiAESA is almost the same than the AESA one. In this work, we report experiments comparing with some competing methods. Our empirical results show that this new approach obtains a significant reduction of distance computations with no execution time penalty.

A Tabular Pruning Rule in Tree-Based Fast Nearest Neighbor Search Algorithms

Some fast nearest neighbor search (NNS) algorithms using metric properties have appeared in the last years for reducing computational cost. Depending on the structure used to store the training set, different strategies to speed up the search have been defined. For instance, pruning rules avoid the search of some branches of a tree in a tree-based search algorithm. In this paper, we propose a new and simple pruning rule that can be used in most of the tree-based search algorithms. All the information needed by the rule can be stored in a table (at preprocessing time). Moreover, the rule can be computed in constant time. This approach is evaluated through real and artificial data experiments. In order to test its performance, the rule is compared to and combined with other previously defined rules.