The AmTriangle meta-dataset for experimenting with Supervised Machine Learning

Authors

DOI:

https://doi.org/10.25746/ruiips.v9.i4.26211

Keywords:

AmTriangle, dataset, Python code, KNN, supervised learning

Abstract

This article presents the first version of a meta-dataset named "AmTriangle", and related tools, designed for introducing and experimenting with supervised machine learning.

It is a "meta-dataset", because it is an instrument for generating sets of samples (datasets), as many as desired, correctly classified. Each sample is a triangle, classified as "acute", "obtuse" or "right", according to trigonometry.

A generated dataset can be used to "teach-by-example" how to classify new samples, by different techniques, such as KNN or neural networks; the results depend on options, discussed. The article makes parallels with other datasets: the "triangles" could be other analogous objects.

Capa: O meta-dataset AmTriangle para experiências com Supervised Machine Learning

Published

2021-12-31

How to Cite

Marques, A. (2021). The AmTriangle meta-dataset for experimenting with Supervised Machine Learning. Revista Da UI_IPSantarém, 9(4), 58–68. https://doi.org/10.25746/ruiips.v9.i4.26211

Most read articles by the same author(s)