274 research outputs found
Guided Zoom: Zooming into Network Evidence to Refine Fine-Grained Model Decisions
No full textIn state-of-the-art deep single-label classification models, the top-k (k = 2,3,4,....) accuracy is usually significantly higher than the top-1 accuracy. This is more evident in fine-grained datasets, where differences between classes are quite subtle. Exploiting the information provided in the top k predicted classes boosts the final prediction of a model. We propose Guided Zoom, a novel way in which explainabitity could be used to improve model performance. We do so by making sure the model has "the right reasons" fora prediction. The reason/evidence upon which a deep neural network makes a prediction is defined to be the grounding, in the pixel space, for a specific class conditional probability in the model output. Guided Zoom examines how reasonable the evidence used to make each of the top-k predictions is. Test time evidence is deemed reasonable if it is coherent with evidence used to make similar correct decisions at training time. This leads to better informed predictions. We explore a variety of grounding techniques and study their complementarity for computing evidence. We show that Guided Zoom results in an improvement of a model's classification accuracy and achieves state-of-the-art classification performance on four fine-grained classification datasets
Guided Zoom: Zooming into Network Evidence to Refine Fine-Grained Model Decisions
No full textIn state-of-the-art deep single-label classification models, the top-kk (k=2,3,4, \dots)(k=2,3,4,⋯) accuracy is usually significantly higher than the top-1 accuracy. This is more evident in fine-grained datasets, where differences between classes are quite subtle. Exploiting the information provided in the top kk predicted classes boosts the final prediction of a model. We propose Guided Zoom, a novel way in which explainability could be used to improve model performance. We do so by making sure the model has 'the right reasons' for a prediction. The reason/evidence upon which a deep neural network makes a prediction is defined to be the grounding, in the pixel space, for a specific class conditional probability in the model output. Guided Zoom examines how reasonable the evidence used to make each of the top-kk predictions is. Test time evidence is deemed reasonable if it is coherent with evidence used to make similar correct decisions at training time. This leads to better informed predictions. We explore a variety of grounding techniques and study their complementarity for computing evidence. We show that Guided Zoom results in an improvement of a model's classification accuracy and achieves state-of-the-art classification performance on four fine-grained classification datasets. Our code is available at https://github.com/andreazuna89/Guided-Zoom
Guided Zoom: Questioning Network Evidence for Fine-Grained Classification
No full textWe propose Guided Zoom, an approach that utilizes spatial grounding of a model’s decision to make more informed predictions. It does so by making sure the model has “the right reasons” for a prediction, defined as reasons that are coherent with those used to make similar correct decisions at training time. The reason/evidence upon which a deep convolutional neural network makes a prediction is defined to be the spatial grounding, in the pixel space, for a specific class conditional probability in the model output. Guided Zoom examines how reasonable such evidence is for each of the top-k predicted classes, rather than solely trusting the top-1 prediction. We show that Guided Zoom improves the classification accuracy of a deep convolutional neural network model and obtains state-of-the-art results on three fine-grained classification benchmark datasets
Not All Neuro-Symbolic Concepts Are Created Equal: Analysis and Mitigation of Reasoning Shortcuts
No full textNeuro-Symbolic (NeSy) predictive models hold the promise of improved compliance with given constraints, systematic generalization, and interpretability, as they
allow to infer labels that are consistent with some prior knowledge by reasoning
over high-level concepts extracted from sub-symbolic inputs. It was recently shown
that NeSy predictors are affected by reasoning shortcuts: they can attain high accuracy but by leveraging concepts with unintended semantics, thus coming short of
their promised advantages. Yet, a systematic characterization of reasoning shortcuts
and of potential mitigation strategies is missing. This work fills this gap by characterizing them as unintended optima of the learning objective and identifying four
key conditions behind their occurrence. Based on this, we derive several natural
mitigation strategies, and analyze their efficacy both theoretically and empirically.
Our analysis shows reasoning shortcuts are difficult to deal with, casting doubts on
the trustworthiness and interpretability of existing NeSy solutions
Thermodynamic modeling and experimental investigation of the MgO-Y2O3-ZrO2 system
No full textSolid-state phase equilibria in the MgO-Y2O3-ZrO2 system as well as the equilibria including liquid were investigated in the whole-compositional range using high-temperature differential thermal analysis (DTA), X-ray diffraction (XRD), and scanning electron microscopy combined with energy dispersive X-ray spectroscopy (SEM/EDX). Isothermal sections at 1493, 1573, 1693, and 1923 K were constructed based on experimental studies. The presence of tie line between MgO and Y4Zr3O12 in the temperature range between 1493 and 1573 K was confirmed. The eutectic melting in the MgO-Y2O3-ZrO2 system was established using DTA followed by SEM/EDX microstructure investigation. Based on the obtained experimental results, the thermodynamic database was derived
Attacks on Online Learners: a Teacher-Student Analysis
No full textMachine learning models are famously vulnerable to adversarial attacks: small adhoc perturbations of the data that can catastrophically alter the model predictions.
While a large literature has studied the case of test-time attacks on pre-trained
models, the important case of attacks in an online learning setting has received
little attention so far. In this work, we use a control-theoretical perspective to study
the scenario where an attacker may perturb data labels to manipulate the learning
dynamics of an online learner. We perform a theoretical analysis of the problem
in a teacher-student setup, considering different attack strategies, and obtaining
analytical results for the steady state of simple linear learners. These results enable
us to prove that a discontinuous transition in the learner’s accuracy occurs when
the attack strength exceeds a critical threshold. We then study empirically attacks
on learners with complex architectures using real data, confirming the insights of
our theoretical analysis. Our findings show that greedy attacks can be extremely
efficient, especially when data stream in small batche
The Pick-to-Learn Algorithm: Empowering Compression for Tight Generalization Bounds and Improved Post-training Performance
No full text
Fast and simple spectral clustering in theory and practice
No full textSpectral clustering is a popular and effective algorithm designed to find k clusters in a graph G. In the classical spectral clustering algorithm, the vertices of G are embedded into ℝk using k eigenvectors of the graph Laplacian matrix. However, computing this embedding is computationally expensive and dominates the running time of the algorithm. In this paper, we present a simple spectral clustering algorithm based on a vertex embedding with O(log(k)) vectors computed by the power method. The vertex embedding is computed in nearly-linear time with respect to the size of the graph, and the algorithm provably recovers the ground truth clusters under natural assumptions on the input graph. We evaluate the new algorithm on several synthetic and real-world datasets, finding that it is significantly faster than alternative clustering algorithms, while producing results with approximately the same clustering accuracy
- …
