FairALM: Augmented Lagrangian Method for Training Fair Models with Little Regret

@article{Lokhande2020FairALMAL,
  title={FairALM: Augmented Lagrangian Method for Training Fair Models with Little Regret},
  author={Vishnu Suresh Lokhande and Aditya Kumar Akash and Sathya Ravi and Vikas Singh},
  journal={Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision},
  year={2020},
  volume={12357},
  pages={
          365-381
        }
}
  • Vishnu Suresh Lokhande, A. K. Akash, Vikas Singh
  • Published 3 April 2020
  • Computer Science
  • Computer vision - ECCV ... : ... European Conference on Computer Vision : proceedings. European Conference on Computer Vision
Algorithmic decision making based on computer vision and machine learning methods continues to permeate our lives. But issues related to biases of these models and the extent to which they treat certain segments of the population unfairly, have led to legitimate concerns. There is agreement that because of biases in the datasets we present to the models, a fairness-oblivious training will lead to unfair models. An interesting topic is the study of mechanisms via which the de novo design or… 

F3: Fair and Federated Face Attribute Classification with Heterogeneous Data

TLDR
F3 adopts multiple heuristics to improve fairness across different demographic groups without requiring data homogeneity assumption, and suggests that F3 strikes a practical balance between accuracy and fairness for FAC.

Fair Federated Learning for Heterogeneous Data

TLDR
This work considers the problem of achieving fair classification in Federated Learning under data heterogeneity, and proposes several aggregation techniques that are empirically validated by comparing the resulting fairness and accuracy on CelebA and UTK datasets.

Are My Deep Learning Systems Fair? An Empirical Study of Fixed-Seed Training

TLDR
This paper conducts the first empirical study to quantify the impact of software implementation on the fairness and its variance of DL systems and calls for better fairness evaluation and testing protocols to improve fairness and fairness variance ofDL systems as well as DL research validity and reproducibility at large.

Fair Federated Learning for Heterogeneous Face Data

TLDR
This work considers the problem of achieving fair classification in Federated Learning under data heterogeneity, and proposes several aggregation techniques that are empirically validated by comparing the resulting fairness metrics and accuracy on CelebA, UTK, and FairFace datasets.

A Survey of Fairness in Medical Image Analysis: Concepts, Algorithms, Evaluations, and Challenges

Fairness, a criterion focuses on evaluating algorithm performance on di ff erent demographic groups, has gained attention in natural language processing, recommendation system and facial recognition.

Through a fair looking-glass: mitigating bias in image datasets

TLDR
This study presents a fast and effective model to de-bias an image dataset through reconstruction and minimizing the statistical dependence between intended variables, and achieves a promising fairness-accuracy combination.

On the Versatile Uses of Partial Distance Correlation in Deep Learning

TLDR
This paper revisits a (less widely known) from statistics, called distance correlation (and its partial variant), designed to evaluate correlation between feature spaces of different dimensions, and suggests a versatile regularizer with many advantages, which avoids some of the common difficulties one faces in such analyses.

Learning Tucker Compression for Deep CNN

TLDR
Experiments show that LTC can make a network like ResNet, VGG faster with nearly the same classification accuracy, which surpasses current tensor decomposition approaches.

Competitive Physics Informed Networks

TLDR
Numerical experiments show that a CPINN trained with competitive gradient descent can achieve errors two orders of magnitude smaller than that of a PINNtrained with Adam or stochastic gradient descent.

Group-Aware Threshold Adaptation for Fair Classification

TLDR
A novel post-processing method to optimize over multiple fairness constraints through group-aware threshold adaptation to learn adaptive classification thresholds for each demographic group by optimizing the confusion matrix estimated from the probability distribution of a classification model output.

References

SHOWING 1-10 OF 51 REFERENCES

Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints

TLDR
This work proposes to inject corpus-level constraints for calibrating existing structured prediction models and design an algorithm based on Lagrangian relaxation for collective inference to reduce the magnitude of bias amplification in multilabel object classification and visual semantic role labeling.

A Reductions Approach to Fair Classification

TLDR
The key idea is to reduce fair classification to a sequence of cost-sensitive classification problems, whose solutions yield a randomized classifier with the lowest (empirical) error subject to the desired constraints.

Two-Player Games for Efficient Non-Convex Constrained Optimization

TLDR
It is proved that this proxy-Lagrangian formulation, instead of having unbounded size, can be taken to be a distribution over no more than m+1 models (where m is the number of constraints), which is a significant improvement in practical terms.

Situation Recognition: Visual Semantic Role Labeling for Image Understanding

This paper introduces situation recognition, the problem of producing a concise summary of the situation an image depicts including: (1) the main activity (e.g., clipping), (2) the participating

Learning Deep Features for Discriminative Localization

In this work, we revisit the global average pooling layer proposed in [13], and shed light on how it explicitly enables the convolutional neural network (CNN) to have remarkable localization ability

Proximal Algorithms

TLDR
The many different interpretations of proximal operators and algorithms are discussed, their connections to many other topics in optimization and applied mathematics are described, some popular algorithms are surveyed, and a large number of examples of proxiesimal operators that commonly arise in practice are provided.

Dissecting racial bias in an algorithm used to manage the health of populations

TLDR
It is suggested that the choice of convenient, seemingly effective proxies for ground truth can be an important source of algorithmic bias in many contexts.

Competitive Gradient Descent

We introduce a new algorithm for the numerical computation of Nash equilibria of competitive two-player games. Our method is a natural generalization of gradient descent to the two-player setting

Critical reflection on professional development in the social sciences: interview results

Purpose – The aim of this paper is to present an interview and postscript that examine the specific meaning, rationale, conceptual framework, assessment and teaching of critical reflection in and on

Constrained Optimization and Lagrange Multiplier Methods

...