Search

When classifiers face imbalanced class distributions, they often misclassify minority class samples, consequently diminishing the predictive performance of machine learning models. Existing oversampling techniques predominantly rely on the selection of neighboring data via interpolation, with less emphasis on uncovering the intrinsic patterns and relationships within the data. In this research, we present the usefulness of an algorithm named RuLer to deal with the problem of classification with imbalanced data. RuLer is a learning algorithm initially designed to recognize new sound patterns within the context of the performative artistic practice known as live coding. This paper demonstrates that this algorithm, once adapted (Ad-RuLer), has great potential to address the problem of oversampling imbalanced data. An extensive comparison with other mainstream oversampling algorithms (SMOTE, ADASYN, Tomek-links, Borderline-SMOTE, and KmeansSMOTE), using different classifiers (logistic regression, random forest, and XGBoost) is performed on several real-world datasets with different degrees of data imbalance. The experiment results indicate that Ad-RuLer serves as an effective oversampling technique with extensive applicability.

In this paper, an e-Learning toolbox based on a set of fuzzy logic data mining techniques is presented. The toolbox is mainly based on the fuzzy inductive reasoning (FIR) methodology and two of its key extensions: (i) the linguistic rules extraction algorithm (LR-FIR), which extracts comprehensible and consistent sets of rules describing students’ learning behavior, and (ii) the causal relevance approach (CR-FIR), which allows to reduce uncertainty during a student’s performance prediction stage, and provides a relative weighting of the features involved in the evaluation process. In addition, the presented toolbox enables, in an incremental way, detecting and grouping students with respect to their learning behavior, with the main goal to timely detect failing students, and properly provide them with suitable and actionable feedback. The proposed toolbox has been applied to two different datasets gathered from two courses at the Latin American Institute for Educational Communication virtual campus. The introductory and didactic planning courses were analyzed using the proposed toolbox. The results obtained by the functionalities offered by the platform allow teachers to make decisions and carry out improvement actions in the current course, i.e., to monitor specific student clusters, to analyze possible changes in the different evaluable activities, or to reduce (to some extent) teacher workload.

This manuscript explores fuzzy rule learning for sound synthesizer programming within the performative practice known as live coding. In this practice, sound synthesis algorithms are programmed in real time by means of source code. To facilitate this, one possibility is to automatically create variations out of a few synthesizer presets. However, the need for real-time feedback makes existent synthesizer programmers unfeasible to use. In addition, sometimes presets are created mid-performance and as such no benchmarks exist. Inductive rule learning has shown to be effective for creating real-time variations in such a scenario. However, logical IF-THEN rules do not cover the whole feature space. Here, we present an algorithm that extends IF-THEN rules to hyperrectangles, which are used as the cores of membership functions to create a map of the input space. To generalize the rules, the contradictions are solved by a maximum volume heuristics. The user controls the novelty-consistency balance with respect to the input data using the algorithm parameters. The algorithm was evaluated in live performances and by cross-validation using extrinsic-benchmarks and a dataset collected during user tests. The model’s accuracy achieves state-of-the-art results. This, together with the positive criticism received from live coders that tested our methodology, suggests that this is a promising approach.

The usage of indicators as constituent parts of composite indices is an extended practice in many fields of knowledge. Even if rigorous statistical analyses are implemented, many of the methodologies follow simple arithmetic assumptions to aggregate indicators to build an index. One of the consequences of such assumptions can be the concealment of the influence of some of the composite index’s components. We developed a fuzzy method that aggregates indicators using non-linear methods and, in this paper, compare it to a well-known example in the field of risk assessment, called Moncho’s equation, which combines physical and social components and uses a linear aggregation method to estimate a level of seismic risk. By comparing the spatial pattern of the risk level obtained from these two methodologies, we were able to evaluate to what extent a fuzzy approach allows a more realistic representation of how social vulnerability levels might shape the seismic risk panorama in an urban environment. We found that, in some cases, this approach can lead to risk level values that are up to 80% greater than those obtained using a linear aggregation method for the same areas.

The energy consumption used for domestic purposes in Europe is, to a considerable extent, due to heating and cooling. This energy is produced mostly by burning fossil fuels, which has a high negative environmental impact. The characteristics of a building are an important factor to determine the necessities of heating and cooling loads. Therefore, the study of the relevant characteristics of the buildings, regarding the heating and cooling needed to maintain comfortable indoor air conditions, could be very useful in order to design and construct energy-efficient buildings. In previous studies, different machine-learning approaches have been used to predict heating and cooling loads from the set of variables: relative compactness, surface area, wall area, roof area, overall height, orientation, glazing area and glazing area distribution. However, none of these methods are based on fuzzy logic. In this research, we study two fuzzy logic approaches, i.e., fuzzy inductive reasoning (FIR) and adaptive neuro fuzzy inference system (ANFIS), to deal with the same problem. Fuzzy approaches obtain very good results, outperforming all the methods described in previous studies except one. In this work, we also study the feature selection process of FIR methodology as a pre-processing tool to select the more relevant variables before the use of any predictive modelling methodology. It is proven that FIR feature selection provides interesting insights into the main building variables causally related to heating and cooling loads. This allows better decision making and design strategies, since accurate cooling and heating load estimations and correct identification of parameters that affect building energy demands are of high importance to optimize building designs and equipment specifications.