Improving Feature Attribution through Input-specific Network Pruning

Khakzar, Ashkan; Baselizadeh, Soroosh; Khanduja, Saurabh; Rupprecht, Christian; Kim, Seong-Tae; Navab, Nassir

doi:10.48550/arxiv.1911.11081

Cited by 4 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the perturbation-based XAI methods have the challenge of combinatorial complexity explosion. This happens when one attempts to go through all elements of the input and all their possible combinations to observe how each of them would affect the output [62]. The possible combinations of data perturbations increase dramatically when dealing with 3D images, causing a significant increase in computational costs.…”

Section: B Explainable Methodsmentioning

confidence: 99%

sMRI-PatchNet: A Novel Efficient Explainable Patch-Based Deep Learning Network for Alzheimer’s Disease Diagnosis With Structural MRI

Zhang,

Han,

Han

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Structural magnetic resonance imaging (sMRI) can identify subtle brain changes due to its high contrast for soft tissues and high spatial resolution. It has been widely used in diagnosing neurological brain diseases, such as Alzheimer's disease (AD). However, the size of 3D high-resolution data poses a significant challenge for data analysis and processing. Since only a few areas of the brain show structural changes highly associated with AD, the patch-based methods dividing the whole data into several regular patches have shown promising for more efficient image analysis. The major challenges of the patch-based methods include identifying the discriminative patches, combining features from the discrete discriminative patches, and designing appropriate classifiers. This work proposes a novel efficient patchbased deep learning network (sMRI-PatchNet) with explainable patch localisation and selection for AD diagnosis. Specifically, it consists of two primary components: 1) A fast and efficient explainable patch selection method for determining the most discriminative patches; and 2) A novel patch-based network for extracting deep features and AD classification with position embeddings to retain position information, capable of capturing the global and local information of inter-and intra-patches. This method has been applied for the AD classification and the prediction of the transitional state moderate cognitive impairment (MCI) conversion with real datasets. The experimental evaluation shows that the proposed method can identify discriminative pathological locations effectively with a significant reduction on patch numbers used, providing better performance in terms of accuracy, computing performance, and generalizability, in contrast to the state-of-the-art methods.

show abstract

Section: B Explainable Methodsmentioning

confidence: 99%

sMRI-PatchNet: A Novel Efficient Explainable Patch-Based Deep Learning Network for Alzheimer’s Disease Diagnosis With Structural MRI

Zhang,

Han,

Han

et al. 2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…If it were to iterate over all input elements and all possible combinations of them and observe how each one changes the output, this would be impractical in practice due to enormous cost constraints. Gradient-based attribution methods calculate the importance of elements by using model gradients, but tend to generate noise, especially in particularly large networks [13]. In addition, if the backpropagation algorithm is used, it needs to access the internal information of the model to generate the explanation, so it will be limited in use.…”

Section: Perturbation-based Saliency Methodsmentioning

confidence: 99%

Research on Reinforcement Learning Explainable Strategies Based on Advantage Saliency

Yan¹

2023

FCIS

View full text Add to dashboard Cite

Deep reinforcement learning is increasingly being used in difficult environments with sparse rewards and high-dimensional inputs, and it performs well, but its decision-making processes are largely unclear and difficult to explain to end users. Saliency map methods explain an agent's behavior by highlighting state features relevant for the agent to take an action. In this paper, we use the perturbation-based saliency map method, propose the use of advantage function to replace the existing method of calculating state saliency, realize the combination of advantage function and perturbation-based saliency map. A saliency map is generated by noting the saliency of the dependent elements of the agent's chosen action in the Atari game environment. Experimental comparisons show that our method generates more accurate explanatory saliency maps.

show abstract

“…Others address this problem by multi-task learning [10,26] that used selective instances for training on imbalanced sets for each task. Interpreting Neural Networks: Two principal neural network interpretation approaches are feature attribution [29,32,13,30,22,17,12] (i.e. saliency methods [4]) and analyzing internal units (e.g.…”

Section: Related Workmentioning

confidence: 99%

Analyzing the Effects of Handling Data Imbalance on Learned Features from Medical Images by Looking Into the Models

Khakzar¹,

Li²,

Zhang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

One challenging property lurking in medical datasets is the imbalanced data distribution, where the frequency of the samples between the different classes is not balanced. Training a model on an imbalanced dataset can introduce unique challenges to the learning problem where a model is biased towards the highly frequent class. Many methods are proposed to tackle the distributional differences and the imbalanced problem. However, the impact of these approaches on the learned features is not well studied. In this paper, we look deeper into the internal units of neural networks to observe how handling data imbalance affects the learned features. We study several popular cost-sensitive approaches for handling data imbalance and analyze the feature maps of the convolutional neural networks from multiple perspectives: analyzing the alignment of salient features with pathologies and analyzing the pathology-related concepts encoded by the networks. Our study reveals differences and insights regarding the trained models that are not reflected by quantitative metrics such as AUROC and AP and show up only by looking at the models through a lens.

show abstract

Improving Feature Attribution through Input-specific Network Pruning

Cited by 4 publications

References 0 publications

sMRI-PatchNet: A Novel Efficient Explainable Patch-Based Deep Learning Network for Alzheimer’s Disease Diagnosis With Structural MRI

sMRI-PatchNet: A Novel Efficient Explainable Patch-Based Deep Learning Network for Alzheimer’s Disease Diagnosis With Structural MRI

Research on Reinforcement Learning Explainable Strategies Based on Advantage Saliency

Analyzing the Effects of Handling Data Imbalance on Learned Features from Medical Images by Looking Into the Models

Contact Info

Product

Resources

About