Analysing the Fairness of Fairness-aware Classifiers
نویسندگان
چکیده
Calders and Verwer’s two-naive-Bayes is one of fairness-aware classifiers, which classify objects while excluding the influence of a specific information. We analyze why this classifier achieves very high level of the fairness, and show that this is due to a decision rules and a model bias. Based on these findings, we develop methods that are grounded on rigid theory and are applicable to wider types of classifiers.
منابع مشابه
A New Fairness Index and Novel Approach for QoS-Aware Resource Allocation in LTE Networks Based on Utility Functions
Resource allocation techniques have recently appeared as a widely recognized feature in LTE networks. Most of existing approaches in resource allocation focus on maximizing network’s utility functions. The great potential of utility function in improving resource allocation and enhancing fairness and mean opinion score (MOS) indexes has attracted large efforts over the last few years. In this p...
متن کاملOn Fairness, Diversity and Randomness in Algorithmic Decision Making
Consider a binary decision making process where a single machine learning classifier replaces a multitude of humans. We raise questions about the resulting loss of diversity in the decision making process. We study the potential benefits of using random classifier ensembles instead of a single classifier in the context of fairness-aware learning and demonstrate various attractive properties: (i...
متن کاملCross-layer Packet-dependant OFDM Scheduling Based on Proportional Fairness
This paper assumes each user has more than one queue, derives a new packet-dependant proportional fairness power allocation pattern based on the sum of weight capacity and the packet’s priority in users’ queues, and proposes 4 new cross-layer packet-dependant OFDM scheduling schemes based on proportional fairness for heterogeneous classes of traffic. Scenario 1, scenario 2 and scenario 3 lead r...
متن کاملThe Effect of Personality Traits and Equity Sensitivity on the Employees’ Fairness Perception in the Lebanese Pharmaceutical Field
The pharmaceutical sector plays a fundamental role in supporting the economic development of a country. The performance of this sector is influenced by employees’ performance, which can be affected by the way employees perceive fairness at their workplace. Fairness perception can be affected by the equity sensitivity type and the personality traits of employees. The Lebanese context lacks studi...
متن کاملMultisided Fairness for Recommendation
Recent work on machine learning has begun to consider issues of fairness. In this paper, we extend the concept of fairness to recommendation. In particular, we show that in some recommendation contexts, fairness may be a multisided concept, in which fair outcomes for multiple individuals need to be considered. Based on these considerations, we present a taxonomy of classes of fairness-aware rec...
متن کامل