Abstract
This article serves as an introduction to the Special Issue on Metalearning and Algorithm Selection. The introduction is divided into two parts. In the the first section, we give an overview of how the field of metalearning has evolved in the last 1-2 decades and mention how some of the papers in this special issue fit in. In the second section, we discuss the contents of this special issue. We divide the papers into thematic subgroups, provide information about each subgroup, as well as about the individual papers. Our main aim is to highlight how the papers selected for this special issue contribute to the field of metalearning.

Abstract
Selecting appropriate classification algorithms for a given dataset is crucial and useful in practice but is also full of challenges. In order to maximize performance, users of machine learning algorithms need methods that can help them identify the most relevant features in datasets, select algorithms and determine their appropriate hyperparameter settings. In this paper, a method of recommending classification algorithms is proposed. It is oriented towards the average ranking method, combining algorithm rankings observed on prior datasets to identify the best algorithms for a new dataset. Our method uses a special case of data mining workflow that combines algorithm selection preceded by a feature selection method (CFS).

Abstract

Abstract

Abstract
The agribusiness volatility is related to the uncertainty of the environment, rising demand, falling prices and new technologies. However, generation of agriculture data has increased over past years and can be used for a growing number of applications of data mining techniques in agriculture. The multidisciplinary approach of integrating computer science with agriculture will support the necessary decisions to be taken in order to mitigate risks and maximize profits. The present study analyzes different methods of regression applied in the study case of grapes production forecast. The selected methods were multivariate linear regression, regression trees, lasso and random forest. Their performance were compared against the predictions obtained by the company through the mean squared error and the coefficient of variation. The four regression methods used obtained better predictive results than the method used by the company with statistical significance < 0.5%. © Springer International Publishing AG, part of Springer Nature 2018.

Abstract
The moment immediately before the "add to cart" decision is very critical in online shopping. Drawing on theories of transfer, spreading activation and human-computer interaction, the superiority of markerless Augmented Reality (AR) and Marker-based augmented reality (M) over Conventional Interactive (CI) is hypothesized. Although those multimedia tools are not part of the product/brand motivating the consumer interest they interfere in the interactive performance of the ecommerce. 150 consumers in a lab experiment showed higher emotional response, interactive response and brand evaluation in M and AR than CI. Contrary to what was expected the usability results were the inverse. That is, usability of CI outperforms M and AR. Considering only AR and M interfaces their effect on psychological variables was not statistically significant. A sophisticated or a simple interface had no impact on intention to buy the target brand, but the brand recommendation improved from M to AR. The differing effect of those three interface systems was mediated by brand familiarity, perceived risk, opinion leadership and positive emotional traits.