Visual Customer Segmentation and Data Exploration with the Self-Organizing Map

Advancements in information technology have made it possible for organizations to gather and store vast amounts of data of their customers. Information stored in databases can be highly valuable for organizations. However, analyzing large databases has proven to be difficult in practice. For companies in the retail industry, customer intelligence can be used to identify profitable customers, their characteristics, and behavior. By clustering customers into homogeneous groups, companies can more effectively manage their customer base and target profitable customer segments. This thesis will study the use of the self-organizing map (SOM) as a method for analyzing large customer datasets, clustering customers, and discovering information about customer behavior. Aim of the thesis is to find out whether the SOM could be a practical tool for retail companies to analyze their customer data.

Searching for music archives and composition manuscripts from the Viola database

Music archives and composition manuscripts from the Viola database.

Wood species identification by visual characterization of fibers

Kandidaatintyö tehtiin osana PulpVision-tutkimusprojektia, jonka tarkoituksena on kehittää kuvapohjaisia laskenta- ja luokittelumetodeja sellun laaduntarkkailuun paperin valmistuksessa. Tämän tutkimusprojektin osana on aiemmin kehitetty metodi, jolla etsittiin kaarevia rakenteita kuvista, ja tätä metodia hyödynnettiin kuitujen etsintään kuvista. Tätä metodia käytettiin lähtökohtana kandidaatintyölle. Työn tarkoituksena oli tutkia, voidaanko erilaisista kuitukuvista laskettujen piirteiden avulla tunnistaa kuvassa olevien kuitujen laji. Näissä kuitukuvissa oli kuituja neljästä eri puulajista ja yhdestä kasvista. Nämä lajit olivat akasia, koivu, mänty, eukalyptus ja vehnä. Jokaisesta lajista valittiin 100 kuitukuvaa ja nämä kuvat jaettiin kahteen ryhmään, joista ensimmäistä käytettiin opetusryhmänä ja toista testausryhmänä. Opetusryhmän avulla jokaiselle kuitulajille laskettiin näitä kuvaavia piirteitä, joiden avulla pyrittiin tunnistamaan testausryhmän kuvissa olevat kuitulajit. Nämä kuvat oli tuottanut CEMIS-Oulu (Center for Measurement and Information Systems), joka on mittaustekniikkaan keskittynyt yksikkö Oulun yliopistossa. Yksittäiselle opetusryhmän kuitukuvalle laskettiin keskiarvot ja keskihajonnat kolmesta eri piirteestä, jotka olivat pituus, leveys ja kaarevuus. Lisäksi laskettiin, kuinka monta kuitua kuvasta löydettiin. Näiden piirteiden eri yhdistelmien avulla testattiin tunnistamisen tarkkuutta käyttämällä k:n lähimmän naapurin menetelmää ja Naiivi Bayes -luokitinta testausryhmän kuville. Testeistä saatiin lupaavia tuloksia muun muassa pituuden ja leveyden keskiarvoja käytettäessä saavutettiin jopa noin 98 %:n tarkkuus molemmilla algoritmeilla. Tunnistuksessa kuitujen keskimäärinen pituus vaikutti olevan kuitukuvia parhaiten kuvaava piirre. Käytettyjen algoritmien välillä ei ollut suurta vaihtelua tarkkuudessa. Testeissä saatujen tulosten perusteella voidaan todeta, että kuitukuvien tunnistaminen on mahdollista. Testien perusteella kuitukuvista tarvitsee laskea vain kaksi piirrettä, joilla kuidut voidaan tunnistaa tarkasti. Käytetyt lajittelualgoritmit olivat hyvin yksinkertaisia, mutta ne toimivat testeissä hyvin.

In search of good relationships to music : understanding aspiration and challenge in developing music school teacher practices

This study focuses on teacher practices in publicly funded music schools in Finland. As views on the aims of music education change and broaden, music schools across Europe share the challenge of developing their activities in response. In public and scholarly debate, there have been calls for increased diversity of contents and concepts of teaching. In Finland, the official national curriculum for state-funded music schools builds on the ideal that teaching and learning should create conditions which promote ‘a good relationship to music’. The meaning of this concept has been deliberately left open in order to leave room for dialogue, flexibility, and teacher autonomy. Since what is meant by ‘good’ is not defined in advance, the notion of ‘improving’ practices is also open to discussion. The purpose of the study is to examine these issues from teachers’ point of view by asking what music school teachers aim to accomplish as they develop their practices. Methodologically, the study introduces a suggestion for building empirical research on Alperson’s ‘robust’ praxial approach to music education, a philosophical theory which is strongly committed to practitioner perspectives and musical diversity. A systematic method for analysing music education practices, interpretive practice analysis, is elaborated with support from interpretive research methods originally used in policy analysis. In addition, the research design shows how reflecting conversations (a collaborative approach well-known in Nordic social work) can be fruitfully applied in interpretive research and combined with teacher inquiry. Data have been generated in a collaborative project involving five experienced music school teachers and the researcher. The empirical material includes transcripts from group conversations, data from teacher inquiry conducted within the project, and transcripts from follow-up interviews. The teachers’ aspirations can be understood as strivings to reinforce the connection between musical practices and various forms of human flourishing such that music and flourishing can sustain each other. Examples from their practices show how the word ‘good’ receives its meaning in context. Central among the teachers’ concerns is their hope that students develop a free and sustainable interest in music, often described as inspiration. I propose that ‘good relationships to music’ and ‘inspiration’ can be understood as philosophical mediators which support the transition from an indeterminate ‘interest in music’ towards specific ways in which music can become a (co-)constitutive part of living well in each person’s particular circumstances. Different musical practices emphasise different aspects of what is considered important in music and in human life. Music school teachers consciously balance between a variety of such values. They also make efforts to resist pressure which might threaten the goods they think are most important. Such goods include joy, participation, perseverance, solid musical skills related to specific practices, and a strong sense of vitality. The insights from this study suggest that when teachers are able to create inspiration, they seem to do so by performing complex work which combines musical and educational aims and makes general positive contributions to their students’ lives. Ensuring that teaching and learning in music schools remain as constructive and meaningful as possible for both students and teachers is a demanding task. The study indicates that collaborative, reflective and interdisciplinary work may be helpful as support for development processes on both individual and collective levels of music school teacher practices.

Pretraining Convolutional Neural Networks for Visual Recognition

Convolutional Neural Networks (CNN) have become the state-of-the-art methods on many large scale visual recognition tasks. For a lot of practical applications, CNN architectures have a restrictive requirement: A huge amount of labeled data are needed for training. The idea of generative pretraining is to obtain initial weights of the network by training the network in a completely unsupervised way and then fine-tune the weights for the task at hand using supervised learning. In this thesis, a general introduction to Deep Neural Networks and algorithms are given and these methods are applied to classification tasks of handwritten digits and natural images for developing unsupervised feature learning. The goal of this thesis is to find out if the effect of pretraining is damped by recent practical advances in optimization and regularization of CNN. The experimental results show that pretraining is still a substantial regularizer, however, not a necessary step in training Convolutional Neural Networks with rectified activations. On handwritten digits, the proposed pretraining model achieved a classification accuracy comparable to the state-of-the-art methods.

Publication Statistics Music recordings 1995-

Vapaakappalekartuntaan perustuva tilasto Suomessa julkaistuista musiikkiäänitteistä vuodesta 1995 lähtien

Publication Statistics Sheet Music 1991-

Vapaakappalekartuntaan perustuva tilasto Suomessa kustannetuista nuottijulkaisuista vuodesta 1991 lähtien

La vitrina, herramienta de negocios y promoción visual : Consejos prácticos

Bogotá Emprende

La vitrina: herramienta de negocios y promoción visual : Guía de consulta

Bogotá Emprende