This thesis explores the ideas of feature learning and sparse coding for Music Information Retrieval (MIR). Sparse coding is an algorithm which aims to learn new feature representations from data automatically. In contrast to previous work which uses sparse coding in an MIR context the concept of supervised sparse coding is also investigated, which makes use of the ground-truth labels explicitly during the learning process. Here sparse coding and supervised coding are applied to two MIR problems: classification of musical genre and recognition of the emotional content of music. A variation of Label Consistent K-SVD is used to add supervision during the dictionary learning process. In the case of Music Genre Recognition (MGR) an additional discriminative term is added to encourage tracks from the same genre to have similar sparse codes. For Music Emotion Recognition (MER) a linear regression term is added to learn an optimal classifier and dictionary pair. These results indicate that while sparse coding performs well for MGR, the additional supervision fails to improve the performance. In the case of MER, supervised coding significantly outperforms both standard sparse coding and commonly used designed features, namely MFCC and pitch chroma.

Nowadays, outlier detection has attracted growing interest. Unlike typical outlier detection problems, relational outlier detection focuses on detecting abnormal patterns in datasets that contain relational implications within each data point. Furthermore, different from the traditional outlier detection that focuses on only numerical data, modern outlier detection models must be able to handle data in various types and structures. Detecting relational outliers should consider (1) Dependencies among different data types, (2) Data types that are not continuous or do not have ordinal characteristics, such as binary, categorical or multi-label, and (3) Special structures in the data. This thesis focuses on the development of relational outlier detection methods and real-world applications in datasets that contain non-numerical, mixed-type, and special structure data in three tasks, namely (1) outlier detection in mixed-type data, (2) categorical outlier detection in music genre data, and (3) outlier detection in categorized time series data. For the first task, existing solutions for mixed-type data mostly focus on computational efficiency, and their strategies are mostly heuristic driven, lacking a statistical foundation. The proposed contributions of our work include: (1) Constructing a novel unsupervised framework based on a robust generalized linear model (GLM), (2) Developing a model that is capable of capturing large variances of outliers and dependencies among mixed-type observations, and designing an approach for approximating the analytically intractable Bayesian inference, and (3) Conducting extensive experiments to validate effectiveness and efficiency. For the second task, we extended and applied the modeling strategy to a real-world problem. The existing solutions to the specific task are mostly supervised, and the traditional outlier detection methods only focus on detecting outliers by the data distributions, ignoring the input-output relation between the genres and the extracted features. The proposed contributions of our work for this task include: (1) Proposing an unsupervised outlier detection framework for music genre data, (2) Extending the GLM based model in the first task to handle categorical responses and developing an approach to approximate the analytically intractable Bayesian inference, and (3) Conducting experiments to demonstrate that the proposed method outperforms the benchmark methods. For the third task, we focused on improving the outlier detection performance in the second task by proposing a novel framework and expanded the research scope to general categorized time-series data. Existing studies have suggested a large number of methods for automatic time series classification. However, there is a lack of research focusing on detecting outliers from manually categorized time series. The proposed contributions of our work for this task include: (1) Proposing a novel semi-supervised robust outlier detection framework for categorized time-series datasets, (2) Further extending the new framework to an active learning system that takes user insights into account, and (3) Conducting a comprehensive set of experiments to demonstrate the performance of the proposed method in real-world applications. / Doctor of Philosophy / In recent years, outlier detection has been one of the most important topics in the data mining and machine learning research domain. Unlike typical outlier detection problems, relational outlier detection focuses on detecting abnormal patterns in datasets that contain relational implications within each data point. Detecting relational outliers should consider (1) Dependencies among different data types, (2) Data types that are not continuous or do not have ordinal characteristics, such as binary, categorical or multi-label, and (3) Special structures in the data. This thesis focuses on the development of relational outlier detection methods and real-world applications in datasets that contain non-numerical, mixed-type, and special structure data in three tasks, namely (1) outlier detection in mixed-type data, (2) categorical outlier detection in music genre data, and (3) outlier detection in categorized time series data. The first task aims on constructing a novel unsupervised framework, developing a model that is capable of capturing the normal pattern and the effects, and designing an approach for model fitting. In the second task, we further extended and applied the modeling strategy to a real-world problem in the music technology domain. For the third task, we expanded the research scope from the previous task to general categorized time-series data, and focused on improving the outlier detection performance by proposing a novel semi-supervised framework.

This diploma work deals with music genre recognition using the techniques of Music Information Retrieval. It contains a brief description of the principle of this research area and its subfield called Music Genre Recognition. The following chapter includes selection of the most suitable parameters for describing music genres. This work further characterizes machine learning methods used in this field of research. The next chapter deals with the descriptions of music datasets created for genre classification studies. Subsequently, there is a draft and evaluation of the system for music genre recognition. The last part of this work describes the results of partial parameter analysis, dependence of genre classification accuracy on the amount of parameters and contains a discussion on the causes of classification accurancy for the individual genres.

A teoria das redes complexas tem se tornado cada vez mais em uma poderosa teoria computacional capaz de representar, caracterizar e examinar sistemas com estrutura não trivial, revelando características intrínsecas locais e globais que facilitam a compreensão do comportamento e da dinâmica de tais sistemas. Nesta tese são exploradas as vantagens das redes complexas na resolução de problemas relacionados com tarefas do âmbito musical, especificamente, são estudadas três abordagens: reconhecimento de padrões, mineração e síntese de músicas. A primeira abordagem é desempenhada através do desenvolvimento de um método para a extração do padrão rítmico de uma peça musical de caráter popular. Nesse tipo de peças coexistem diferentes espécies de padrões rítmicos, os quais configuram uma hierarquia que é determinada por aspectos funcionais dentro da base rítmica. Os padrões rítmicos principais são caracterizados por sua maior incidência dentro do discurso musical, propriedade que é refletida na formação de comunidades dentro da rede. Técnicas de detecção de comunidades são aplicadas na extração dos padrões rítmicos, e uma medida para diferenciar os padrões principais dos secundários é proposta. Os resultados mostram que a qualidade da extração é sensível ao algoritmo de detecção, ao modo de representação do ritmo e ao tratamento dado às linhas de percussão na hora de gerar a rede. Uma fase de mineração foi desempenhada usando medidas topológicas sobre a rede obtida após a remoção dos padrões secundários. Técnicas de aprendizado supervisionado e não-supervisionado foram aplicadas para discriminar o gênero musical segundo os atributos calculados na fase de mineração. Os resultados revelam a eficiência da metodologia proposta, a qual foi constatada através de um teste de significância estatística. A última abordagem foi tratada mediante o desenvolvimento de modelos para a composição de melodias através de duas perspectivas, na primeira perspectiva é usada uma caminhada controlada por critérios sobre redes complexas predefinidas e na segunda redes neurais recorrentes e sistemas dinâmicos caóticos. Nesta última perspectiva, o modelo é treinado para compor uma melodia com um valor preestabelecido de alguma característica tonal subjetiva através de uma estratégia de controle proporcional que modifica a complexidade de uma melodia caótica, melodia que atua como entrada de inspiração da rede. / The theory of complex networks has become increasingly a powerful computational tool capable of representing, characterizing and examining systems with non-trivial structure, revealing both local and global intrinsic structures that facilitate the understanding of the behavior and dynamics of such systems. In this thesis, the virtues of complex networks in solving problems related to tasks within the musical scope are explored. Specifically, three approaches are studied: pattern recognition, data mining, and synthesis. The first perspective is addressed by developing a method for extracting the rhythmic pattern of a piece of popular music. In that type of musical pieces, there coexist different types of rhythm patterns which constitute a hierarchy determined by functional aspects within the basic rhythm. The main rhythmic patterns are characterized by a higher incidence within the musical discourse and this factor is reflected in the formation of communities within the network constructed from the music piece. Community detection techniques are applied in the extraction of rhythmic patterns, and a measure to distinguish the main patterns of the secondary is proposed. The results showed that the quality of extraction is sensitive to the detection algorithm, the method of representing rhythm, and treatment of percussion lines when generating the network. Data mining is performed using topological measures over the network obtained after the removal of secondary patterns. Techniques of supervised and unsupervised learning are applied to discriminate the musical genre according to the attributes calculated in the data mining phase. The quantitative results show the efficiency of the proposed methodology, which is confirmed by a test of statistical significance. Regarding the melody generation, an algorithm using a walk controlled by criteria on predefined complex networks has been developed, as well as the development of melody composition models using recurrent neural networks and chaotic dynamical systems. In the last approach, the model is trained to compose a melody with a subjective characteristic melodic value pre-established by a proportional control strategy that acts on the parameters of a chaotic melody as input inspiration.

A teoria das redes complexas tem se tornado cada vez mais em uma poderosa teoria computacional capaz de representar, caracterizar e examinar sistemas com estrutura não trivial, revelando características intrínsecas locais e globais que facilitam a compreensão do comportamento e da dinâmica de tais sistemas. Nesta tese são exploradas as vantagens das redes complexas na resolução de problemas relacionados com tarefas do âmbito musical, especificamente, são estudadas três abordagens: reconhecimento de padrões, mineração e síntese de músicas. A primeira abordagem é desempenhada através do desenvolvimento de um método para a extração do padrão rítmico de uma peça musical de caráter popular. Nesse tipo de peças coexistem diferentes espécies de padrões rítmicos, os quais configuram uma hierarquia que é determinada por aspectos funcionais dentro da base rítmica. Os padrões rítmicos principais são caracterizados por sua maior incidência dentro do discurso musical, propriedade que é refletida na formação de comunidades dentro da rede. Técnicas de detecção de comunidades são aplicadas na extração dos padrões rítmicos, e uma medida para diferenciar os padrões principais dos secundários é proposta. Os resultados mostram que a qualidade da extração é sensível ao algoritmo de detecção, ao modo de representação do ritmo e ao tratamento dado às linhas de percussão na hora de gerar a rede. Uma fase de mineração foi desempenhada usando medidas topológicas sobre a rede obtida após a remoção dos padrões secundários. Técnicas de aprendizado supervisionado e não-supervisionado foram aplicadas para discriminar o gênero musical segundo os atributos calculados na fase de mineração. Os resultados revelam a eficiência da metodologia proposta, a qual foi constatada através de um teste de significância estatística. A última abordagem foi tratada mediante o desenvolvimento de modelos para a composição de melodias através de duas perspectivas, na primeira perspectiva é usada uma caminhada controlada por critérios sobre redes complexas predefinidas e na segunda redes neurais recorrentes e sistemas dinâmicos caóticos. Nesta última perspectiva, o modelo é treinado para compor uma melodia com um valor preestabelecido de alguma característica tonal subjetiva através de uma estratégia de controle proporcional que modifica a complexidade de uma melodia caótica, melodia que atua como entrada de inspiração da rede. / The theory of complex networks has become increasingly a powerful computational tool capable of representing, characterizing and examining systems with non-trivial structure, revealing both local and global intrinsic structures that facilitate the understanding of the behavior and dynamics of such systems. In this thesis, the virtues of complex networks in solving problems related to tasks within the musical scope are explored. Specifically, three approaches are studied: pattern recognition, data mining, and synthesis. The first perspective is addressed by developing a method for extracting the rhythmic pattern of a piece of popular music. In that type of musical pieces, there coexist different types of rhythm patterns which constitute a hierarchy determined by functional aspects within the basic rhythm. The main rhythmic patterns are characterized by a higher incidence within the musical discourse and this factor is reflected in the formation of communities within the network constructed from the music piece. Community detection techniques are applied in the extraction of rhythmic patterns, and a measure to distinguish the main patterns of the secondary is proposed. The results showed that the quality of extraction is sensitive to the detection algorithm, the method of representing rhythm, and treatment of percussion lines when generating the network. Data mining is performed using topological measures over the network obtained after the removal of secondary patterns. Techniques of supervised and unsupervised learning are applied to discriminate the musical genre according to the attributes calculated in the data mining phase. The quantitative results show the efficiency of the proposed methodology, which is confirmed by a test of statistical significance. Regarding the melody generation, an algorithm using a walk controlled by criteria on predefined complex networks has been developed, as well as the development of melody composition models using recurrent neural networks and chaotic dynamical systems. In the last approach, the model is trained to compose a melody with a subjective characteristic melodic value pre-established by a proportional control strategy that acts on the parameters of a chaotic melody as input inspiration.