• No results found

Median Variants of Prototype Based Learning Vector Quantization: Methods for Classification of General Proximity Data

N/A
N/A
Info
Downloaden
Protected

Academic year: 2021

Share "Median Variants of Prototype Based Learning Vector Quantization: Methods for Classification of General Proximity Data"

Copied!
2
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Download het nu ( 2 pagina )

Hele tekst

(1)

University of Groningen

Median Variants of Prototype Based Learning Vector Quantization

Nebel, David

DOI:

10.33612/diss.135377546

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2020

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Nebel, D. (2020). Median Variants of Prototype Based Learning Vector Quantization: Methods for

Classification of General Proximity Data. University of Groningen. https://doi.org/10.33612/diss.135377546

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

Stellingen

behorende bij het proefschrift

Median Variants of Prototype Based Learning

Vector Quantization Methods for Classification

in Case of General Proximity Data

van

David Nebel

1. Precise knowledge of the properties of a given proximity measure is necessary for a good machine learning model.

2. Similarities are not necessarily inner products/kernels and vice versa. 3. Different proximity measures induce different neighbourhood relations

between data objects. A respective mathematical analysis before neural network training helps to avoid pitfalls.

4. Pre-processing of proximity data may drastically change neighbourhood relations in an undesired manner.

5. Expectation Maximization is a very powerful optimization technique also for non-probabilistic optimization problems.

6. Median algorithms provide sparse and interpretable models, which satisfy at least a good lower bound for the classification accuracy. 7. A set of proximity measures is like a group of humans. If one proximity

seems less helpful, a group of proximities can still benefit from this one. 8. Sometimes research is a strong fight between competing experts.

However, it can be very beneficial for both fighters.

Referenties

Download het nu ( PDF - 2 pagina - 609.59 KB )

GERELATEERDE DOCUMENTEN

Gekkenwerk, maar wel leuk gekkenwerk

Want niet alleen wij waren aan het werk in de tuin: tegelijkertijd maakte Teleac opnames voor een program- ma over vogelvriendelijk tuinieren.. Te zien

Verslag: archeologisch proefonderzoek Ezelstraat, Brugge

De eerste sleuf bevindt zich op de parking langs de Jan Boninstraat en is 16,20 m lang, de tweede op de parking langs de Hugo Losschaertstraat is 8 m lang.. Dit pakket bestaat

Kernel-based data fusion for machine learning: methods and applications in bioinformatics and text mining

Moreover, we solidify our data sources and algorithms in a gene prioritization software, which is characterized as a novel kernel-based approach to combine text mining data

RISK-BASED DATA CLASSIFICATION

Data owners need to be assigned and users need to be identified because; these roles are key in the identification and valuation of information assets, they can impose

Learning Vector Quantization for Ordinal Classification: p-OGMLVQ

In p-OGMLVQ, the correct and incorrect prototype classes are identified by the label distance from the true class using an absolute error loss function and a ranking loss threshold..

GPLVQ Prototype-based Probabilistic Classification

Many machine learning algorithms build their classification model by iterating over the input and comparing the calculated training results to the training data.. A Gaussian process

Finding order in the noise : a conceptual overview on the shift to complex dynamical models in psychology

The most important shift in the function of complex dynamical models is that the focus is on the system’s process over time, not on the underlying, linear relationship

Discursieve strategieën van Islamitische Staat

(2009) onder andere onderzoek naar discursieve strategieën die worden gebruikt voor de constructie van de Oostenrijkse nationale identiteit, terwijl Küçükalî (2015)