作者:
Szymon BobekGrzegorz J. NalepaFaculty of Physics
Astronomy and Applied Computer Science Institute of Applied Computer Science Department of Human-Centered Artificial Intelligence ul. prof. Stanisława Łojasiewicza 11 Krakow 30-348 Poland
Deep neural networks (DNNs) are highly effective at extracting features from complex data types, such as images and text, but often function as black-box models, making interpretation difficult. We propose TSProto — ...
详细信息
Deep neural networks (DNNs) are highly effective at extracting features from complex data types, such as images and text, but often function as black-box models, making interpretation difficult. We propose TSProto — a model-agnostic approach that goes beyond standard XAI methods focused on feature importance, clustering important segments into conceptual prototypes — high-level, human-interpretable units. This approach not only enhances transparency but also avoids issues seen with surrogate models, such as the Rashomon effect, enabling more direct insights into DNN behavior. Our method involves two phases: (1) using feature attribution tools (e.g., SHAP, LIME) to highlight regions of model importance, and (2) fusion of these regions into prototypes with contextual information to form meaningful concepts. These concepts then integrate into an interpretable decision tree, making DNNs more accessible for expert analysis. We benchmark our solution on 61 publicly available datasets, where it outperforms other state-of-the-art prototype-based methods and glassbox models by an average of 10% in the F1 metric. Additionally, we demonstrate its practical applicability in a real-life anomaly detection case. The results from the user evaluation, conducted with 17 experts recruited from leading European research teams and industrial partners, also indicate a positive reception among experts in XAI and the industry. Our implementation is available as an open-source Python package on GitHub and PyPi.
The model of incomplete cooperative games incorporates uncertainty into the classical model of cooperative games by considering a partial characteristic function. Thus the values for some of the coalitions are not kno...
详细信息
The model of incomplete cooperative games incorporates uncertainty into the classical model of cooperative games by considering a partial characteristic function. Thus the values for some of the coalitions are not known. The main focus of this paper is the class of 1-convex cooperative games under this framework. We are interested in two heavily intertwined questions. First, given an incomplete game, in which ways can we fill in the missing values to obtain a classical 1-convex game? Such complete games are called 1-convex extensions. For the class of minimal incomplete games (in which precisely the values of singletons and grand coalitions are known), we provide an answer in terms of a description of the set of 1-convex extensions. The description employs extreme points and extreme rays of the set. We also provide bounds on sets of 1-convex extensions for such games. Second, how to determine in a rational, fair, and efficient way the payoffs of players based only on the known values of coalitions? Based on the description of the set of 1-convex extensions, we introduce generalisations of three solution concepts (values) for complete games, namely the τ-value, the Shapley value and the nucleolus. We consider two variants where we compute the centre of gravity of either extreme games or of a combination of extreme games and extreme rays. We show that all of the generalised values coincide for minimal incomplete games which allows to introduce the average value. For this value, we provide three different axiomatisations based on axiomatic characterisations of the τ-value and the Shapley value for classical cooperative games. Finally, we turn our attention to incomplete games with defined upper vector, asking the same questions and this time arriving to different conclusions. This provides a benchmark to test our tools and knowledge of the average value. This part highlights the importance and also the difficulty of considering more general classes of incomplete games.
We revisit the antiferromagnetic structure of Tb14Ag51 (Fischer et al., 2005) with the propagation vector [ 1313, 0] and parent space group P6/m using both magnetic symmetry and irreducible representation arguments. W...
详细信息
Spatial distributions of wisents (European bison) Bison bonasus (Linnaeus 1758), were studied in the Bieszczady Mountains (south-eastern Poland) on the basis of telemetric data obtained from two subpopulations: the we...
详细信息
ISBN:
(数字)9798331530631
ISBN:
(纸本)9798331530648
Spatial distributions of wisents (European bison) Bison bonasus (Linnaeus 1758), were studied in the Bieszczady Mountains (south-eastern Poland) on the basis of telemetric data obtained from two subpopulations: the western – inhabiting the forest districts of Baligród, Komańcza, Cisna and Lesko, and the eastern – dwelling within the districts of Lutowiska, Stuposiany, and in Bieszczadzki National park. Data was collected between 2002-2021. In our study, we propose a novel approach for classifying wisent subpopulations that utilizes machine learning methods and CLC data. For this purpose, the performances of eight algorithms: Naive Bayes, Logistic Regression, Support Vector Machine, Multilayer Perceptron, Random Forest, Extreme Gradient Boosting, k-Nearest Neighbors and Decision Tree were investigated. The algorithms were compared according to the following indicators: accuracy, Cohen’s-Kappa, precision, recall and F1 score. Their assessment was enhanced through the application of statistical inference (the Friedman test with post-hoc analysis) and SHAP values. The lowest results were achieved by Naive Bayes and Logistic Regression methods (accuracy of 73.69% and 74.43%, respectively), whilst significantly higher results were achieved by eXtreme Gradient Boosting, k-Nearest Neighbors and Decision Tree classifiers, with the classification accuracy exceeding 90% (91.81%, 92.8%, 93.47%, respectively). Based on the results, we conclude that CLC data represents a valuable source of information regarding the affinity of wisents towards different habitat conditions.
We determine thresholds pc for random-site percolation on a triangular lattice for all available neighborhoods containing sites from the first to the fifth coordination zones, including their complex combinations. The...
详细信息
The dynamics of social relations and the possibility of reaching the state of structural balance (Heider balance) are discussed for various networks of interacting actors under the influence of the temperature modelin...
详细信息
Protein nanoparticles have been proven to be highly effective stabilizers of water-in-water emulsions obtained from a number of different types of aqueous two-phase systems (ATPS). The emulsion stabilizing efficiency ...
详细信息
The wounded-nucleon and -quark models are compared using d+Au collisions at sNN = 200 GeV. The shape of the wounded-quark emission function seems to be universal for different centralities, in contrast to the wounded-...
详细信息
We present results on the Central Exclusive Production of charged particle pairs h+h− (h = π, K, p), pp → p + h+h− + p, obtained with the STAR experiment at RHIC in proton-proton collisions at a center-of-mass energ...
详细信息
We study a simple deterministic map that leads a fully connected network to Heider balance. The map is realized by an algorithm that updates all links synchronously in a way depending on the state of the entire networ...
暂无评论