اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدDirectional Bias Amplification
118
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Mitigating bias in machine learning systems requires refining our understanding of bias propagation pathways: from societal structures to large-scale data to trained models to impact on society. In this work, we focus on one aspect of the problem, namely bias amplification: the tendency of models to amplify the biases present in the data they are trained on. A metric for measuring bias amplification was introduced in the seminal work by Zhao et al. (2017); however, as we demonstrate, this metric suffers from a number of shortcomings including conflating different types of bias amplification and failing to account for varying base rates of protected attributes. We introduce and analyze a new, decoupled metric for measuring bias amplification, $text{BiasAmp}_{rightarrow}$ (Directional Bias Amplification). We thoroughly analyze and discuss both the technical assumptions and normative implications of this metric. We provide suggestions about its measurement by cautioning against predicting sensitive attributes, encouraging the use of confidence intervals due to fluctuations in the fairness of models across runs, and discussing the limitations of what this metric captures. Throughout this paper, we work to provide an interrogative look at the technical measurement of bias amplification, guided by our normative ideas of what we want it to encompass. Code is located at https://github.com/princetonvisualai/directional-bias-amp
قيم البحث
اقرأ أيضاً
Recommender systems usually amplify the biases in the data. The model learned from historical interactions with imbalanced item distribution will amplify the imbalance by over-recommending items from the major groups. Addressing this issue is essenti
al for a healthy ecosystem of recommendation in the long run. Existing works apply bias control to the ranking targets (e.g., calibration, fairness, and diversity), but ignore the true reason for bias amplification and trade-off the recommendation accuracy. In this work, we scrutinize the cause-effect factors for bias amplification, identifying the main reason lies in the confounder effect of imbalanced item distribution on user representation and prediction score. The existence of such confounder pushes us to go beyond merely modeling the conditional probability and embrace the causal modeling for recommendation. Towards this end, we propose a Deconfounded Recommender System (DecRS), which models the causal effect of user representation on the prediction score. The key to eliminating the impact of the confounder lies in backdoor adjustment, which is however difficult to do due to the infinite sample space of the confounder. For this challenge, we contribute an approximation operator for backdoor adjustment which can be easily plugged into most recommender models. Lastly, we devise an inference strategy to dynamically regulate backdoor adjustment according to user status. We instantiate DecRS on two representative models FM and NFM, and conduct extensive experiments over two benchmarks to validate the superiority of our proposed DecRS.
The popularity and widespread use of pruning and quantization is driven by the severe resource constraints of deploying deep neural networks to environments with strict latency, memory and energy requirements. These techniques achieve high levels of
compression with negligible impact on top-line metrics (top-1 and top-5 accuracy). However, overall accuracy hides disproportionately high errors on a small subset of examples; we call this subset Compression Identified Exemplars (CIE). We further establish that for CIE examples, compression amplifies existing algorithmic bias. Pruning disproportionately impacts performance on underrepresented features, which often coincides with considerations of fairness. Given that CIE is a relatively small subset but a great contributor of error in the model, we propose its use as a human-in-the-loop auditing tool to surface a tractable subset of the dataset for further inspection or annotation by a domain expert. We provide qualitative and quantitative support that CIE surfaces the most challenging examples in the data distribution for human-in-the-loop auditing.
Dropout is a simple but effective technique for learning in neural networks and other settings. A sound theoretical understanding of dropout is needed to determine when dropout should be applied and how to use it most effectively. In this paper we co
ntinue the exploration of dropout as a regularizer pioneered by Wager, et.al. We focus on linear classification where a convex proxy to the misclassification loss (i.e. the logistic loss used in logistic regression) is minimized. We show: (a) when the dropout-regularized criterion has a unique minimizer, (b) when the dropout-regularization penalty goes to infinity with the weights, and when it remains bounded, (c) that the dropout regularization can be non-monotonic as individual weights increase from 0, and (d) that the dropout regularization penalty may not be convex. This last point is particularly surprising because the combination of dropout regularization with any convex loss proxy is always a convex function. In order to contrast dropout regularization with $L_2$ regularization, we formalize the notion of when different sources are more compatible with different regularizers. We then exhibit distributions that are provably more compatible with dropout regularization than $L_2$ regularization, and vice versa. These sources provide additional insight into how the inductive biases of dropout and $L_2$ regularization differ. We provide some similar results for $L_1$ regularization.
Algorithmic approaches endow deep learning systems with implicit bias that helps them generalize even in over-parametrized settings. In this paper, we focus on understanding such a bias induced in learning through dropout, a popular technique to avoi
d overfitting in deep learning. For single hidden-layer linear neural networks, we show that dropout tends to make the norm of incoming/outgoing weight vectors of all the hidden nodes equal. In addition, we provide a complete characterization of the optimization landscape induced by dropout.
Building reliable machine learning systems requires that we correctly understand their level of confidence. Calibration measures the degree of accuracy in a models confidence and most research in calibration focuses on techniques to improve an empiri
cal estimate of calibration error, ECE_bin. We introduce a simulation framework that allows us to empirically show that ECE_bin can systematically underestimate or overestimate the true calibration error depending on the nature of model miscalibration, the size of the evaluation data set, and the number of bins. Critically, we find that ECE_bin is more strongly biased for perfectly calibrated models. We propose a simple alternative calibration error metric, ECE_sweep, in which the number of bins is chosen to be as large as possible while preserving monotonicity in the calibration function. Evaluating our measure on distributions fit to neural network confidence scores on CIFAR-10, CIFAR-100, and ImageNet, we show that ECE_sweep produces a less biased estimator of calibration error and therefore should be used by any researcher wishing to evaluate the calibration of models trained on similar datasets.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
