Table of Contents

1. Introduction
   1.1. Scope and Overview
   1.2. Notation and Conventions
2. Related Work
   2.1. Content-Based Music Retrieval
   2.2. Approaches using Self-Organizing Maps
3. Feature Extraction
   3.1. Raw Audio Data
   3.2. Loudness Sensation
        3.2.1. Discrete Fourier Transformation
        3.2.2. Critical-Bands
        3.2.3. Masking
        3.2.4. Decibel
        3.2.5. Phon
        3.2.6. Sone
        3.2.7. Examples
   3.3. Dynamics
        3.3.1. Amplitude Modulated Loudness
        3.3.2. Fluctuation Strength
        3.3.3. Modified Fluctuation Strength
        3.3.4. Examples
   3.4. Computational Efficiency
   3.5. Summary
4. Clustering
   4.1. Self-Organizing Maps
        4.1.1. Background
        4.1.2. The Batch SOM Algorithm
   4.2. Music Sequences
   4.3. Pieces of Music
        4.3.1. Principal Component Analysis
        4.3.2. Gaussian Mixture Models
        4.3.3. Fuzzy C-Means
        4.3.4. K-Means
        4.3.5. Median
  4.9. Summary
5. Visualization and User Interface
   5.1. Islands
        5.1.1. Alternatives
   5.2. Labeling
        5.2.1. Aggregated Attributes
        5.2.2. Rhythm
        5.2.3. Results
        5.2.4. HTML Interface
        5.2.5. Summary
6. Conclusion
   6.1. Summary
   6.2. Future Work
A. Source Code and Constants
   A.1. Source Code
        A.1.1. Batch SOM
        A.1.2. Fuzzy C-Means
        A.1.3. K-Means
   A.2. Constants
        A.2.1. Equal Loudness Contours
        A.2.2. Modified Fluctuation Strength
        A.2.3. Color Scale
B. Music Collection