| fullrecord |
[{"key": "dc.contributor.advisor", "value": "Honkanen, Risto", "language": "", "element": "contributor", "qualifier": "advisor", "schema": "dc"}, {"key": "dc.contributor.advisor", "value": "Hakala, Ismo", "language": "", "element": "contributor", "qualifier": "advisor", "schema": "dc"}, {"key": "dc.contributor.author", "value": "Haasiom\u00e4ki, Mika-Petteri", "language": "", "element": "contributor", "qualifier": "author", "schema": "dc"}, {"key": "dc.date.accessioned", "value": "2019-11-29T11:10:01Z", "language": null, "element": "date", "qualifier": "accessioned", "schema": "dc"}, {"key": "dc.date.available", "value": "2019-11-29T11:10:01Z", "language": null, "element": "date", "qualifier": "available", "schema": "dc"}, {"key": "dc.date.issued", "value": "2019", "language": "", "element": "date", "qualifier": "issued", "schema": "dc"}, {"key": "dc.identifier.uri", "value": "https://jyx.jyu.fi/handle/123456789/66582", "language": null, "element": "identifier", "qualifier": "uri", "schema": "dc"}, {"key": "dc.description.abstract", "value": "\u00c4\u00e4nihavainnon pohjalta suoritettavaa automaattista luokitusta voidaan hy\u00f6dynt\u00e4\u00e4 populaation kehityksen seurannassa tai kiinnostavan lajin tunnistamisessa. Luokittelijan kehitt\u00e4minen voi olla vaativaa, joten ty\u00f6ss\u00e4 k\u00e4sitell\u00e4\u00e4n koneoppimisen k\u00e4ytt\u00e4mist\u00e4 luokittelijan kehityksess\u00e4, keskittyen neuroverkkomenetelmiin. Neuroverkot ovat yksi koneoppimisen menetelm\u00e4, jossa sy\u00f6te kuvataan tulokseksi viem\u00e4ll\u00e4 se verkon laskentayksik\u00f6iden l\u00e4pi. Tutkimuskysymyksen\u00e4 on laatia t\u00e4m\u00e4 luokittelija ja tutkia kuinka sen s\u00e4\u00e4dett\u00e4v\u00e4t hyperparametrit vaikuttavat luokittelutarkkuuteen.\n\nTeoriaosuus koostuu katsauksesta koko luokitusprosessin elementtien teoriaan. Osuudessa k\u00e4yd\u00e4\u00e4n l\u00e4pi \u00e4\u00e4nisignaalista koostuvan aineiston k\u00e4sittely, segmentointi ja kiinnostavia tapahtumia kuvaavien piirteiden irrotus. Seuraavaksi k\u00e4yd\u00e4\u00e4n l\u00e4pi neuroverkon elementtien teoria, yleisesti k\u00e4ytetyt virhe- ja aktivointifunktiot. Teoriaosuuden loppuosa koostuu neuroverkon opetusprosessin k\u00e4sittelyst\u00e4, sen haasteista ja opetusvaiheen optimointimenetelmist\u00e4. Normalisointimenetelmien k\u00e4sittelyss\u00e4 on painotettu uusimpia menetelmi\u00e4 kuten ryhm\u00e4normalisointia. \n\nTutkimuskysymyksiin haetaan vastausta kokeellisesti viidell\u00e4 testill\u00e4. Empiirinen osassa kuvataan tutkimuksessa toteutettu ymp\u00e4rist\u00f6 ja k\u00e4ytett\u00e4v\u00e4t luokittelijamallit, sek\u00e4 k\u00e4ytett\u00e4v\u00e4 aineisto. Aineiston pohjalta suoritetaan viisi testitapausta, joilla pyrit\u00e4\u00e4n selvitt\u00e4m\u00e4\u00e4n kuinka neuroverkkomalli kannattaa m\u00e4\u00e4ritell\u00e4, kun tavoitteena on minimoida resurssitarve s\u00e4ilytt\u00e4en hyv\u00e4ksytt\u00e4v\u00e4 luokitustarkkuus. Mallia verrattiin l\u00e4himm\u00e4n naapurin menetelm\u00e4\u00e4n perustuvaan luokittelijaan.\nLuokittelutarkkuuden ja F-mitan tulokset osoittavat, ett\u00e4 neuroverkko on tarkempi kuin verrokki l\u00e4himm\u00e4n naapurin menetelm\u00e4 luokittelija. Tulokset vahvistavat my\u00f6s ryhm\u00e4normalisoinnin merkityst\u00e4 ja soveltuvuutta neuroverkon opetukseen. K\u00e4ytt\u00e4m\u00e4ll\u00e4 ryhm\u00e4normalisointia malli oppi nopeammin ja luokitteli tarkemmin kuin dropout-normalisointia k\u00e4ytett\u00e4ess\u00e4.", "language": "fi", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Automatic classification based on sound event can be used to track changes in a animal population or to detect specific species in monitoring area. Other uses include reducing needless wireless transmissions in sensing or monitoring networks. An automatic classifier makes a decision to attach a class label by using function, that maps input features to a class label. Development of accurate classifying function may be difficult, therefore in this thesis we aim to use machine learning, focusing on neural networks, to reach this goal. Neural networks are used in machine learning to map from input to output by flowing data through layered network of computational units.\n\nIn this thesis we take a look at elements of classification process, such as data set handling, noise rejection and segmentation and feature extraction from audio signal. In following chapters, we describe elements of neural networks, common activation and loss functions, training process and associated challenges, as well as regularization and optimization methods used in current networks. Main research question is to implement classifier using neural networks and test impact of various parameters on classification accuracy.\n\nEmpirical section describes used data set, test cases, environment and implementations. Five tests were conducted with focus to determine parameters for a lightweight neural network, while retaining acceptable classification accuracy. Found model was tested against nearest neighbor classifier, which had access to whole training data set during classification, using 10-fold cross-validation. We found that neural network classifier performed better than nearest neighbor based system with regards to classification accuracy and F-measure score. Additionally results enforce previous results where group normalization yields higher accuracy while converging faster compared to dropout normalization. Our results agree with others on effectiveness of group normalization.", "language": "en", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Submitted by Paivi Vuorio (paelvuor@jyu.fi) on 2019-11-29T11:10:01Z\nNo. of bitstreams: 0", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Made available in DSpace on 2019-11-29T11:10:01Z (GMT). No. of bitstreams: 0\n Previous issue date: 2019", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.format.extent", "value": "93", "language": "", "element": "format", "qualifier": "extent", "schema": "dc"}, {"key": "dc.format.mimetype", "value": "application/pdf", "language": null, "element": "format", "qualifier": "mimetype", "schema": "dc"}, {"key": "dc.language.iso", "value": "fin", "language": null, "element": "language", "qualifier": "iso", "schema": "dc"}, {"key": "dc.rights", "value": "In Copyright", "language": "en", "element": "rights", "qualifier": null, "schema": "dc"}, {"key": "dc.subject.other", "value": "piirreirrotus", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "\u00e4\u00e4nen luokittelu", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.title", "value": "\u00c4\u00e4nien luokitteleminen neuroverkoilla", "language": "", "element": "title", "qualifier": null, "schema": "dc"}, {"key": "dc.type", "value": "master thesis", "language": null, "element": "type", "qualifier": null, "schema": "dc"}, {"key": "dc.identifier.urn", "value": "URN:NBN:fi:jyu-201911295070", "language": "", "element": "identifier", "qualifier": "urn", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Pro gradu -tutkielma", "language": "fi", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Master\u2019s thesis", "language": "en", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Informaatioteknologian tiedekunta", "language": "fi", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Faculty of Information Technology", "language": "en", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Informaatioteknologia", "language": "fi", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Information Technology", "language": "en", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "Jyv\u00e4skyl\u00e4n yliopisto", "language": "fi", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "University of Jyv\u00e4skyl\u00e4", "language": "en", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Tietotekniikka", "language": "fi", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Mathematical Information Technology", "language": "en", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "yvv.contractresearch.funding", "value": "0", "language": "", "element": "contractresearch", "qualifier": "funding", "schema": "yvv"}, {"key": "dc.type.coar", "value": "http://purl.org/coar/resource_type/c_bdcc", "language": null, "element": "type", "qualifier": "coar", "schema": "dc"}, {"key": "dc.rights.accesslevel", "value": "openAccess", "language": null, "element": "rights", "qualifier": "accesslevel", "schema": "dc"}, {"key": "dc.type.publication", "value": "masterThesis", "language": null, "element": "type", "qualifier": "publication", "schema": "dc"}, {"key": "dc.subject.oppiainekoodi", "value": "602", "language": "", "element": "subject", "qualifier": "oppiainekoodi", "schema": "dc"}, {"key": "dc.subject.yso", "value": "el\u00e4inten \u00e4\u00e4net", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "neuroverkot", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "koneoppiminen", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.format.content", "value": "fulltext", "language": null, "element": "format", "qualifier": "content", "schema": "dc"}, {"key": "dc.rights.url", "value": "https://rightsstatements.org/page/InC/1.0/", "language": null, "element": "rights", "qualifier": "url", "schema": "dc"}, {"key": "dc.type.okm", "value": "G2", "language": null, "element": "type", "qualifier": "okm", "schema": "dc"}]
|