Author ' s personal copy A Fast Quartet tree heuristic for hierarchical clustering

By Grgoire Montavon, Sebastian Lapuschkin, Alexander Binder, Wojciech Samek, K. Müller. 2010

Subfields: Algorithms & Complexity, Computational Biology & Bioinformatics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Algorithms & Complexity

In Computational Biology & Bioinformatics

Citations in the corpus

SELFEXPLAIN: A Self-Explaining Architecture for Neural Text Classifiers (2021, background): “Approaches that assign relative importance scores to input features through activation differences (Shrikumar et al., 2017; Montavon et al., 2017) are widely adopted for interpretability in computer vision applications.”
Visual Entailment: A Novel Task for Fine-Grained Image Understanding (2019, background): “Recent work addresses this requirement either through saliency-map visualizations [61, 77, 49], attention mechanism [75, 79, 51, 14], or other analysis [30, 39, 56, 58].”
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) (2018, background): “Recent works that use decomposition to explain locally include [94] and [95].”
A Survey of Methods for Explaining Black Box Models (2018, methodology): “An evolution of LRP is the Deep Taylor Decomposition (DTD) method illustrated in Reference [78].”
Explainable Artificial Intelligence: Understanding, Visualizing and Interpreting Deep Learning Models (2017, background): “Recent work [26] also shows close relations to Taylor decomposition, which is a general function analysis tool in mathematics.”
Methods for interpreting and understanding deep neural networks (2017, methodology): “The decomposition process can therefore be repeated in the lower layers, until the first layer of the neural network is reached, thus, performing a deep Taylor decomposition [34].”
Explaining Recurrent Neural Network Predictions in Sentiment Analysis (2017, methodology): “We refer the reader to (Montavon et al., 2017b) for an overview on methods for understanding and interpreting deep neural network predictions.”
Learning how to explain neural networks: PatternNet and PatternAttribution (2017, methodology): “These restrictions are similar to those in the saliency map (Simonyan et al., 2014), DCN (Zeiler & Fergus, 2014), GBP (Springenberg et al., 2015), LRP (Bach et al., 2015) and DTD (Montavon et al., 2017).”
Nothing Else Matters: Model-Agnostic Explanations By Identifying Prediction Invariance (2016, background): “Explanations such as sparse linear models [11] (henceforth called linear LIME) or gradients [2, 10] can still exhibit high precision and low effort (which are de-facto requirements, as there is little point in explaining a model if explanations lead to poor understanding or are too complex) even…”
A unified view of gradient-based attribution methods for Deep Neural Networks (2017, background): “Nevertheless, Saliency maps are usually rather noisy [14, 12, 20] and taking the absolute value prevents the detection of positive and negative evidence that might be present in the input.”
RUDDER: Return Decomposition for Delayed Rewards (2018, methodology): “For decomposing a linear g one can use the Taylor decomposition (a linear approximation) of g with respect to the h [3, 83].”
Explainable machine learning in deployment (2019, methodology): “These explanation functionals roughly fall into two categories: perturbation-based techniques [9, 16, 23, 35, 48, 58] and gradientbased techniques [8, 41, 53, 55, 56, 59].”
Evaluating the Visualization of What a Deep Neural Network Has Learned (2015, methodology): “Finally, an additional justification for the way LRP technically operates can be found in [35], where the method is shown for certain choices of parameters to perform a “deep Taylor decomposition” of the neural network function.”
ERASER: A Benchmark to Evaluate Rationalized NLP Models (2019, background): “Most automatic evaluations of faithfulness measure the impact of perturbing or erasing words or tokens identified as important on model output (Arras et al., 2017; Montavon et al., 2017; Serrano and Smith, 2019; Samek et al., 2016; Jain and Wallace, 2019).”