From cdfaf3a8e1f838184e00d3cbc4c3257b4b1b3bef Mon Sep 17 00:00:00 2001 From: Ria Cheruvu Date: Thu, 25 Oct 2018 13:21:27 -0700 Subject: [PATCH 1/2] Adding ethics resources on ML and privacy with summaries I'm adding two resources that approach considerations on algorithmic decision-making and privacy from the perspective of an ML practitioner with the short summaries. --- resources.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/resources.md b/resources.md index ee0d2a0..8586dbc 100644 --- a/resources.md +++ b/resources.md @@ -59,12 +59,16 @@ Al-Saggaf, Y., & Islam, M. Z. (2015). Data Mining and Privacy of Social Network [Data Science Ethical Framework, via UK.gov](https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/524298/Data_science_ethics_framework_v1.0_for_publication__1_.pdf) +[Understanding Bias in Machine Learning -- Jindong Gu and Daniela Oelke](https://mybinder.org/v2/gh/Jindong-Explainable-AI/Bias_in_Machine_Learning/master?filepath=ML_Bias.ipynb) This article explores three ways bias can be introduced to ML algorithms from the perspective of an ML practitioner. Bias plays a crucial role in influencing algorithmic decision-making (e.g. due to an imbalanced dataset, the algorithm can start to form racist stereotypes), which makes it an important topic to consider for the ethics of data science. + ## Civic Hacking Schrock, A. R. (2016) Civic hacking as data activism and advocacy: A history from publicity to open government data. New Media & Society, 18, 581-599. ## Data Privacy * [A Reactive Approach for Use-Based Privacy, Eleanor Birrell Fred B. Schneider, 2017](http://www.cs.cornell.edu/fbs/publications/UBP.avanance.pdf) +* [Riazi, S. M., B. D. Rouhani, and F. Koushanfar, "Deep Learning on Private Data", to appear in IEEE Security and Privacy Magazine, 03/2018.](http://www.aceslab.org/sites/default/files/PPDL.pdf) The advent of machine learning as a service demands the need for privacy; this article discusses methods for privacy-preserving deep learning and inference. + ## Data Breaches * [Testimony to US Congress about data breaches](https://www.troyhunt.com/heres-what-im-telling-us-congress-about-data-breaches/) From 2783063d132b4503bd73f279d20c6b35050edb6b Mon Sep 17 00:00:00 2001 From: Ria Cheruvu Date: Thu, 25 Oct 2018 18:12:05 -0700 Subject: [PATCH 2/2] Solving merge conflicts --- resources.md | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/resources.md b/resources.md index 8586dbc..cc4d173 100644 --- a/resources.md +++ b/resources.md @@ -10,7 +10,7 @@ [Using Ethical Reasoning to Amplify the Reach and Resonance of Professional Codes of Conduct in Training Big Data Scientists](https://link.springer.com/article/10.1007%2Fs11948-014-9613-1) -[https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/524298/Data_science_ethics_framework_v1.0_for_publication__1_.pdf](https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/524298/Data_science_ethics_framework_v1.0_for_publication__1_.pdf) +[Data Science Ethical Framework](https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/524298/Data_science_ethics_framework_v1.0_for_publication__1_.pdf) [Derman, E. Wilmott, P. (2009). The financial modeler’s manifesto](http://www.uio.no/studier/emner/sv/oekonomi/ECON4135/h09/undervisningsmateriale/FinancialModelersManifesto.pdf) @@ -57,8 +57,6 @@ Al-Saggaf, Y., & Islam, M. Z. (2015). Data Mining and Privacy of Social Network [algorithmic hiring: Zeynep Tufekci](https://www.youtube.com/watch?v=i7exygaylmY) -[Data Science Ethical Framework, via UK.gov](https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/524298/Data_science_ethics_framework_v1.0_for_publication__1_.pdf) - [Understanding Bias in Machine Learning -- Jindong Gu and Daniela Oelke](https://mybinder.org/v2/gh/Jindong-Explainable-AI/Bias_in_Machine_Learning/master?filepath=ML_Bias.ipynb) This article explores three ways bias can be introduced to ML algorithms from the perspective of an ML practitioner. Bias plays a crucial role in influencing algorithmic decision-making (e.g. due to an imbalanced dataset, the algorithm can start to form racist stereotypes), which makes it an important topic to consider for the ethics of data science. ## Civic Hacking