Explainable Deep Learning: A Field Guide for the Uninitiated

Ras, Gabriëlle; Xie, Ning; Gerven, Marcel van

doi:10.48550/arxiv.2004.14545

Cited by 6 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fortunately, in parallel with the development of AI, the field of explainable AI (XAI) has recently emerged. XAI does not represent a single recipe for understanding AI decision-making, but is rather a conceptual framework in which many different methods are being developed with different underlying assumptions about what “explainable” means (Ras et al, 2020 ). Nevertheless, there are several general traits —evaluation criteria—for XAI.…”

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

“…The cultural relativity of ethics prevents us from imprinting a universal moral code on a model. The ability to understand whether an AI's decision is consistent with the moral code of the environment in which it operates is therefore a more viable solution (Ras et al, 2020 ).…”

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

“…According to Ras et al ( 2020 ) XAIs can be divided into three groups: Visualization methods, Distillation methods, and Implicit methods. Visualization methods attempt to highlight the aspects of the input data that contribute highly to the output.…”

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

“…Yet another alternative is Layer-wise Relevance Propagation (LRP Bach et al, 2015 ), a method that measures relevance instead of sensitivity. That is, the strength of the connection input or pixel to a given network output (Ras et al, 2020 ).…”

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

See 3 more Smart Citations

Deep learning in neuroimaging data analysis: Applications, challenges, and solutions

Avberšek

Repovš

2022

Front. Neuroimaging

View full text Add to dashboard Cite

Methods for the analysis of neuroimaging data have advanced significantly since the beginning of neuroscience as a scientific discipline. Today, sophisticated statistical procedures allow us to examine complex multivariate patterns, however most of them are still constrained by assuming inherent linearity of neural processes. Here, we discuss a group of machine learning methods, called deep learning, which have drawn much attention in and outside the field of neuroscience in recent years and hold the potential to surpass the mentioned limitations. Firstly, we describe and explain the essential concepts in deep learning: the structure and the computational operations that allow deep models to learn. After that, we move to the most common applications of deep learning in neuroimaging data analysis: prediction of outcome, interpretation of internal representations, generation of synthetic data and segmentation. In the next section we present issues that deep learning poses, which concerns multidimensionality and multimodality of data, overfitting and computational cost, and propose possible solutions. Lastly, we discuss the current reach of DL usage in all the common applications in neuroimaging data analysis, where we consider the promise of multimodality, capability of processing raw data, and advanced visualization strategies. We identify research gaps, such as focusing on a limited number of criterion variables and the lack of a well-defined strategy for choosing architecture and hyperparameters. Furthermore, we talk about the possibility of conducting research with constructs that have been ignored so far or/and moving toward frameworks, such as RDoC, the potential of transfer learning and generation of synthetic data.

show abstract

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

Section: Deep Learning Applications In Neuroimagingmentioning

confidence: 99%

See 2 more Smart Citations

Deep learning in neuroimaging data analysis: Applications, challenges, and solutions

Avberšek

Repovš

2022

Front. Neuroimaging

View full text Add to dashboard Cite

show abstract

“…In image recognition, another family of methods widely used is based on visualization. They express an explanation by highlighting characteristics of the image that objectively influence the output of a DNN [15]. The best known of them, Grad-CAM [16], creates class activation map using the gradients of the DNN's output with respect to the last convolutional layer.…”

Section: Introductionmentioning

confidence: 99%