Jorge Calera-Rubio

Stochastic Inference of Regular Tree Languages

We generalize a former algorithm for regular language identification from stochastic samples to the case of tree languages or, equivalently, string languages where structural information is available. We also describe a method to compute efficiently the relative entropy between the target grammar and the inferred one, useful for the evaluation of the inference.

Style recognition through statistical event models

Abstract The automatic classification of music fragments into style classes is one challenging problem within the music information retrieval (MIR) domain and also for the understanding of music style perception. This has a number of applications, including the indexation and exploration of musical databases. Some technologies employed in text classification can be applied to this problem. The key point here is to establish the music equivalent to the words in texts. A number of works use the combination of intervals and duration ratios for this purpose. In this paper, different statistical text recognition algorithms are applied to style recognition using this kind of melody representation, exploring their performance for different word sizes.

Probabilistic k-Testable Tree Languages

In this paper, we present a natural generalization of k-gram models for tree stochastic languages based on the k-testable class. In this class of models, frequencies are estimated for a probabilistic regular tree grammar wich is bottom-up deterministic. One of the advantages of this approach is that the model can be updated in an incremental fashion. This method is an alternative to costly learning algorithms (as inside-outside-based methods) or algorithms that require larger samples (as many state merging/splitting methods).

Stochastic k-testable Tree Languages and Applications

In this paper, we describe a generalization for tree stochastic languages of the k-gram models. These models are based on the k- testable class, a subclass of the languages recognizable by ascending tree automata. One of the advantages of this approach is that the probabilistic model can be updated in an incremental fashion. Another feature is that backing-off schemes can be defined. As an illustration of their applicability, they have been used to compress tree data files at a better rate than string-based methods.

Computing the relative entropy between regular tree languages

Abstract Stochastic grammars provide a formal background in order to deal with tasks where a random source of structured data is involved. In particular, stochastic tree grammars can be useful if hierarchical relations are established among the elementary components of the data. Grammatical inference methods are often checked with training samples generated by a known grammar which is later compared to the grammar inferred from the sample. One measure of their similarity is given by the relative entropy between both grammars. In this paper, we describe an efficient procedure to compute the relative entropy between two stochastic deterministic regular tree grammars.

Online Symbolic-Sequence Prediction with Discrete-Time Recurrent Neural Networks

This paper studies the use of discrete-time recurrent neural networks for predicting the next symbol in a sequence. The focus is on online prediction, a task much harder than the classical offline grammatical inference with neural networks. The results obtained show that the performance of recurrent networks working online is acceptable when sequences come from finite-state machines or even from some chaotic sources. When predicting texts in human language, however, dynamics seem to be too complex to be correctly learned in real-time by the net. Two algorithms are considered for network training: real-time recurrent learning and the decoupled extended Kalman filter.

Smoothing and compression with stochastic k-testable tree languages

In this paper, we describe some techniques to learn probabilistic k-testable tree models, a generalization of the well-known k-gram models, that can be used to compress or classify structured data. These models are easy to infer from samples and allow for incremental updates. Moreover, as shown here, backing-off schemes can be defined to solve data sparseness, a problem that often arises when using trees to represent the data. These features make them suitable to compress structured data files at a better rate than string-based methods.

Parsing with probabilistic strictly locally testable tree languages

Probabilistic k-testable models (usually known as k-gram models in the case of strings) can be easily identified from samples and allow for smoothing techniques to deal with unseen events during pattern classification. In this paper, we introduce the family of stochastic k-testable tree languages and describe how these models can approximate any stochastic rational tree language. The model is applied to the task of learning a probabilistic k-testable model from a sample of parsed sentences. In particular, a parser for a natural language grammar that incorporates smoothing is shown.

Melodic Identification Using Probabilistic Tree Automata

Abstract Similarity computation is a difficult issue in music information retrieval tasks, because it tries to emulate the special ability that humans show for pattern recognition in general, and particularly in the presence of noisy data. A number of works have addressed the problem of what is the best representation for symbolic music in this context. The tree representation, using rhythm for defining the tree structure and pitch information for leaf and node labelling has proven to be effective in melodic similarity computation. One of the main drawbacks of this approach is that the tree comparison algorithms are of a high time complexity. In this paper, stochastic k-testable tree-models are applied for computing the similarity between two melodies as a probability. The results are compared to those achieved by tree edit distances, showing that k-testable tree-models outperform other reference methods in both recognition rate and efficiency. The case study in this paper is to identify a snippet query among a set of songs stored in symbolic format. For it, the utilized method must be able to deal with inexact queries and with efficiency for scalability issues.

A text categorization approach for music style recognition

The automatic classification of music files into styles is one challenging problem in music information retrieval and for music style perception understanding. It has a number of applications, like the indexation and exploration of musical databases. Some techniques used in text classification can be applied to this problem. The key point is to establish a music equivalent to the words in texts. A number of works use the combination of intervals and duration ratios for music description. In this paper, different statistical text recognition algorithms are applied to style recognition using this kind of melody representation, exploring their performance for different word sizes and statistical models.