While deep learning has achieved significant success across various fields, the lack of transparency in decision-making remains a major challenge. To enhance interpretability, we propose an Inclusion-Exclusion Integral Neural Network (IEINN), a novel neural network model based on a non-linear integral defined with respect to a non-additive measure. This model is particularly suitable for systems where monotonic relationships exist between inputs and outputs. However, training without enforcing monotonicity constraints may lead to overfitting and difficulties in interpreting learned parameters. To address this, we introduce a monotonicity-preserving regularization term to ensure learned parameters remain consistent with non-additive measure theory. Experimental results across multiple datasets demonstrate that our approach enhances both model interpretability and predictive stability.

The useful balance of flexibility and robustness provided by fuzzy integrals for classification and regression tasks has been well established in the fuzzy research community. With the rise of artificial intelligence and neural networks, we have witnessed great advances in data analysis and decision-making, however the issue of explainability has become more prominent in recent times. It is in this regard that the theory of fuzzy integrals and other sophisticated aggregation frameworks have a lot to offer. In this contribution, a neural network architecture is proposed based on the inclusion-exclusion integral, which is defined with respect to a fuzzy measure and a triangular norm. After presenting the model and learning methodology, we provide some applications of the network to real datasets. The key benefit of this approach is in the ability to derive interpretations using the Shapley value and interaction indices, read directly from the model. However, we also highlight the predictive performance, which is comparable with state-of-the-art machine learning techniques. Our study hence contributes to the aims of explainable artificial intelligence, whereby model flexibility and accuracy is achieved without sacrificing interpretability.

In this paper, we propose a method to determine the rules for attractive poses and generate novel attractive poses based on the discovered rules. Given a set of attractive poses with a specific style and another set of unattractive poses, we can obtain the rules to separate those two sets of example poses by creating a decision tree based on the low-level pose features that are computed from the example poses. In this research, we implemented our approach for two kinds of attractive poses, Hero and JoJo standing poses, and successfully discovered the rules of these styles. We also developed a heuristic kinematics-based pose deformation method based on the discovered rules of attractive poses. The rules can be applied to any input pose with any specified scale. Using our system, a creator can interactively design novel poses.

In this paper, formal representations for user preferences will be provided, which also take fairness in the achievement of probably conflicting goals into account. The analysis of the Bottleneck Flow Control (BFC) algorithm for congestion avoidance in traffic networks will serve as a base to specify extensions of relations among preference sets. In particular, these relations are the lexicographic minimum, the maxmin fairness, and the newly introduced ordered-ordered weighted averaging operator. All of them are shown to become, in a relational sense, maximized by the BFC algorithm. Their further expansion by introducing fuzzy fusion operators in their formal definitions than establishes a comprehensive sets of relations. The application of these relations to the Multiple Relation Analysis (MRA) of subjective video quality evaluation data is demonstrated and gives the conclusion that also fairness criterions are present in such evaluations, at least as a secondary preference criterion.