SIGEVO - inequality project
DDM (Drift Detection Method) -> Estimates classifier error and standard deviation. Assumes that the convergence of the classifier decreases as more training examples arrive.If the classifier error still increases then there is a drift. Learning with drift detection https://www.researchgate.net/profile/Joao_Gama3/publication/220974771_Learning_with_Drift_Detection/links/0912f50ace71293aed000000.pdf
EDDM (Early Drift Detection Method) -> is a modification of DDM Early drift detection method https://www.cs.upc.edu/~abifet/EDDM.pdf
Cumulative sum approach (CUSUM) -> detects a change of a given parameter value of a probability distribution and indicates when the change is significant. WAT? Continuous inspection schemes https://www.jstor.org/stable/pdf/2333009.pdf?casa_token=38UW_CbPxWAAAAAA:ymVQZS_EhFbyvyOk9fZlo0XH-olWDuxhEQXAGMhGSMKYu7CA1ogAI1AP2-xQEfIVcgzqpgJKmxS3cCCga5anqoEf9NsUh0qHGsLd25JHWx7uCqbW-O8x
PageHinkley -> A modification of the CUSUM algorithm. The cumulative difference between observed classifier error and its average is taken into consideration. A study on change detection methods https://www.researchgate.net/profile/Raquel_Sebastiao/publication/267231166_A_Study_on_Change_Detection_Methods/links/588a1980a6fdccb538f1ee1c/A-Study-on-Change-Detection-Methods.pdf
ADWIN -> A window of incoming examples grows until identifying a change in the average value inside the window. Then it finds another window. Their split point between the 2 windoes is considered as an indication of concept drift. Learning from time-changing data with adaptive windowing https://www.cs.upc.edu/~GAVALDA/papers/adwin06.pdf
On the (im)possibility of fairness https://arxiv.org/pdf/1609.07236.pdf The paper revolves arround the definition of fairness in algorithmic terms.
Understanding Fair Labor Practices in a Networked Age https://datasociety.net/pubs/fow/FairLabor.pdf Yet another paper that can help frame the problem
AI-assisted recruitment is biased. Here’s how to make it more fair https://www.weforum.org/agenda/2019/05/ai-assisted-recruitment-is-biased-heres-how-to-beat-it/ also good for framing purposes, it mentions several cases and explains how large companies use AI for hiring people (second part of the article).
AI can be sexist and racist — it’s time to make it fair https://www.nature.com/articles/d41586-018-05707-8 this paper can help establishing the relevance of the problem and the dimension.
Economic Models of (Algorithmic) Discrimination http://www.mlandthelaw.org/papers/goodman2.pdf
Fairer machine learning in the real world: Mitigating discrimination without collecting sensitive data https://journals.sagepub.com/doi/full/10.1177/2053951717743530 It shows that the problem persists, even after removing sensitive information
Its a Mans Wikipedia?Assessing Gender Inequality in an Online Encyclopedia https://www.aaai.org/ocs/index.php/ICWSM/ICWSM15/paper/viewFile/10585/10528