ICM: An Intuitive Model Independent and Accurate Certainty Measure for Machine Learning

@inproceedings{Waa2018ICMAI,
  title={ICM: An Intuitive Model Independent and Accurate Certainty Measure for Machine Learning},
  author={Jasper van der Waa and Jurriaan van Diggelen and Mark A. Neerincx and Stephan Raaijmakers},
  booktitle={ICAART},
  year={2018}
}
End-users of machine learning-based systems benefit from measures that quantify the trustworthiness of the underlying models. Measures like accuracy provide for a general sense of model performance, but offer no detailed information on specific model outputs. Probabilistic outputs, on the other hand, express such details, but they are not available for all types of machine learning, and can be heavily influenced by bias and lack of representative training data. Further, they are often difficult… 

Figures from this paper

The Design and Validation of an Intuitive Confidence Measure
TLDR
A lazy learning meta-model that can predict how likely a given decision is correct and is intended to be easy to understand, which is validated in an experiment and shows that each user is unique in its desires for explanations.
Interpretable Machine Learning from Granular Computing Perspective
TLDR
This work conducts a survey of the state of the art in IML and GrC fields to settle the bases of the possible contribution of each other with aims to build more interpretable and accurately ML models.
Assuring the Machine Learning Lifecycle: Desiderata, Methods, and Challenges
TLDR
This paper provides a comprehensive survey of the state-of-the-art in the assurance of ML, i.e. in the generation of evidence that ML is sufficiently safe for its intended use, at different stages of the machine learning lifecycle.
Assuring the Machine Learning Lifecycle
TLDR
This article provides a comprehensive survey of the state of the art in the assurance of ML, i.e., in the generation of evidence that ML is sufficiently safe for its intended use, at different stages of the machine learning lifecycle.
Learning Human Intention for Taskable Agents
  • Computer Science
  • 2019
TLDR
The use of context-dependent task constraints is studied as an approximation to the human’s intention for how the task should be executed, showing the context-dependency of human intention for agent tasking.
Guidance on the Assurance of Machine Learning in Autonomous Systems (AMLAS)
TLDR
This document introduces a methodology for the Assurance of Machine Learning for use in Autonomous Systems (AMLAS), a process for systematically integrating safety assurance into the development of ML components and for generating the evidence base for explicitly justifying the acceptable safety of these components when integrated into autonomous system applications.
L G ] 1 0 M ay 2 01 9 Assuring the Machine Learning Lifecycle : Desiderata , Methods , and Challenges
Interpretation [57] ✔ ★ Generate tests via simulation ✔ ★ ✩ ✩ Verifier of Random Forests [163] ✔ ★ Verification of ML Libraries [152] ✔ ★ Check for unwanted bias [15] ✔ ★ Use synthetic test data
Using Perceptual and Cognitive Explanations for Enhanced Human-Agent Team Performance
TLDR
A perceptual-cognitive explanation (PeCoX) framework for the development of explanations that address both the perceptual and cognitive foundations of an agent’s behavior, distinguishing between explanation generation, communication and reception is proposed.
2 . 3 Implementing requirements using machine learning ( ML ) Practical guidance – cross-domain
  • Computer Science
  • 2020
TLDR
The term machine learning workflow is used to refer to these stages taken together, which comprises the activities concerned with the deployment of ML models within an operational system, alongside components obtained using traditional software and system engineering methods.
...
1
2
...

References

SHOWING 1-10 OF 34 REFERENCES
Feature Importance Measure for Non-linear Learning Algorithms
TLDR
The Measure of Feature Importance (MFI) is general and can be applied to any arbitrary learning machine (including kernel machines and deep learning) and can detect features that by itself are inconspicuous and only impact the prediction function through their interaction with other features.
"Why Should I Trust You?": Explaining the Predictions of Any Classifier
TLDR
LIME is proposed, a novel explanation technique that explains the predictions of any classifier in an interpretable and faithful manner, by learning aninterpretable model locally varound the prediction.
Nothing Else Matters: Model-Agnostic Explanations By Identifying Prediction Invariance
TLDR
This work proposes anchor-LIME (aLIME), a model-agnostic technique that produces high-precision rule-based explanations for which the coverage boundaries are very clear and is compared to linear LIME with simulated experiments, and demonstrates the flexibility of aLIME with qualitative examples from a variety of domains and tasks.
Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
TLDR
This work takes convolutional neural networks trained to perform well on either the ImageNet or MNIST datasets and finds images with evolutionary algorithms or gradient ascent that DNNs label with high confidence as belonging to each dataset class, and produces fooling images, which are then used to raise questions about the generality of DNN computer vision.
Algorithmic Transparency via Quantitative Input Influence: Theory and Experiments with Learning Systems
TLDR
The transparency-privacy tradeoff is explored and it is proved that a number of useful transparency reports can be made differentially private with very little addition of noise.
Adam: A Method for Stochastic Optimization
TLDR
This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.
A Review and Empirical Evaluation of Feature Weighting Methods for a Class of Lazy Learning Algorithms
TLDR
A class of weight-setting methods for lazy learning algorithms which use performance feedback to assign weight settings demonstrated three advantages over other methods: they require less pre-processing, perform better in the presence of interacting features, and generally require less training data to learn good settings.
Machine Learning in Action
TLDR
You'll use the flexible Python programming language to build programs that implement algorithms for data classification, forecasting, recommendations, and higher-level features like summarization and simplification in code you can reuse.
Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization
TLDR
This work proposes a technique for producing ‘visual explanations’ for decisions from a large class of Convolutional Neural Network (CNN)-based models, making them more transparent and explainable, and shows that even non-attention based models learn to localize discriminative regions of input image.
Weight Uncertainty in Neural Networks
TLDR
This work introduces a new, efficient, principled and backpropagation-compatible algorithm for learning a probability distribution on the weights of a neural network, called Bayes by Backprop, and shows how the learnt uncertainty in the weights can be used to improve generalisation in non-linear regression problems.
...
1
2
3
4
...