From 5f4171d567f2704ebaf1602babe84cce60a302f6 Mon Sep 17 00:00:00 2001 From: Sarah M Brown Date: Tue, 27 Aug 2024 12:16:58 -0400 Subject: [PATCH] recent work and fall project info start --- code.md | 2 -- index.md | 2 +- join.md | 16 ++++++++++++---- projects.md | 16 +++++++++------- 4 files changed, 22 insertions(+), 14 deletions(-) diff --git a/code.md b/code.md index 719bc9c..7bc682b 100644 --- a/code.md +++ b/code.md @@ -16,7 +16,6 @@ Machine learning data Simulators +++ [{fas}`house` ](https://ml4sts.com/ml-sim) - [{fab}`github` ](https://github.com/ml4sts/ml-sim) ::: @@ -32,7 +31,6 @@ Simpson's paradox inspired Fairness Forensics in collaboration with [the OU Data [{fas}`house` ](https://fairnessforensics.github.io/wiggum) - [{fab}`github` ](https://github.com/fairnessforensics/wiggum) ::: diff --git a/index.md b/index.md index c5fe0de..c75ad05 100644 --- a/index.md +++ b/index.md @@ -1,6 +1,6 @@ # ML4STS lab -We study how ML works for science, technology, and society. This means we apply data science using ML models to understand the world with other scientists, study and evaluate those ML systems directly and build tools to help them intervene. Most of our work right now focuses on building more fair AI and some includes collaboration with other scientists. +We study how to do Machine Learning for Socio-technical Systems. This means we apply data science using ML models to understand the world with other scientists, study and evaluate ML systems directly, and build tools to help data scientists to employ best practices. Most of our work right now focuses on fairness of automated decision making systems and some includes collaboration with other scientists. ## News diff --git a/join.md b/join.md index 811f5b3..5816c5e 100644 --- a/join.md +++ b/join.md @@ -1,6 +1,6 @@ # Join the lab - +To join the lab, you must be interested and motivated to taking a socio-technical approach to doing ML research and building ML-powered systems. ## Current URI Undergraduates @@ -12,6 +12,9 @@ If you are interested, please read the website to learn about the work and [sub Typically, I prefer to work with students who have taken CSC310 (with any professor) or CSC392/CSC311 (with me). ### Current Open Projects + +Undergraduates will almost always spend at least one semester on one of these before starting their own project. The most common exception is students with a double major who have a second research advisor and want to do data science too. + ::::{card-carousel} 2 :::{card} Robust ML evaluation Developer tools @@ -74,18 +77,23 @@ My current graduate course is [ML for science and society](https://ml4scisoc.git Please read the website to learn about the work and [submit this form just for current URI students](https://forms.gle/u8JCvdRJaYLbc4To7). The form notifies Dr. Brown by e-mail and allows you to book a meeting as an interview. -New gradutate students may start out working on one of the projects listed above +New graduate students may start out working on one of the projects listed above before taking leadership on a project of their own. ## Prospective Graduate Students +```{tip} +I will look for the lab name in your statment when I read them +``` + Graduate students in our lab are in the URI Computer Science MS or PhD program. If you apply to this program, Dr. Brown will see your application. To be considered to join the lab, mention the lab name and *why* you want to do research aligned with the lab's research goals in your personal statement. + URI CS Graduate Programs do not use GRE scores, if you send an e-mail that includes them it will not be considered. Dr. Brown gets many inquiries from prospective students. To make your e-mail stick out, be sure to mention how your interested align with the labs goals and make it easy to find. To make your e-mail easy to find, use the subject, `Prospective ML4STS Lab member - ` with the appropriate degree based on what you are applying to selected and send your e-mail to `brownsarahm+ml4sts@uri.edu`. Other inquiries may not be addressed. -Dr. Brown priortiizes research integrity and thoughtfulness in research, using automation inappropriately in the process of searching for graduate school is inconsistent with this. Any students who send emails that seem automated will not be considered for lab membership. +Dr. Brown priortiizes research integrity and thoughtfulness in research. Using automation inappropriately in the process of searching for graduate school is inconsistent with demonstrating that you will conduct research with integrity ad a member of the lab. Any students who send emails that seem inappropriately automated will not be considered for lab membership. ## Other inquiries -Currently, Dr. Brown mostly does not have capacity to supervise people outside of the roles above. If you are interested in double checking, however, send an e-mail to `brownsarahm+ml4sts@uri.edu` +Currently, Dr. Brown mostly does not have capacity to supervise people outside of the roles above. If you are interested in double checking, however, send an e-mail to `brownsarahm+ml4sts@uri.edu`. diff --git a/projects.md b/projects.md index 59ac82a..fe953f9 100644 --- a/projects.md +++ b/projects.md @@ -2,7 +2,7 @@ -:::::{grid} 1 2 2 2 +:::::{grid} 2 :gutter: 2 @@ -10,11 +10,11 @@ :img-top: _static/img/fair_dynamics.png - Dynamics of Fairness +Dynamics of Fairness ^^^ - +(inactive, seeking a student) How does fairness generalize, tolerate distribution shift, and propagate through interacting systems? +++ @@ -28,7 +28,7 @@ How does fairness generalize, tolerate distribution shift, and propagate through :img-top: _static/img/statistical_perceptions.png -Fairness Perceptions +Perceptions of AI Fairness ^^^ In collaboration with [Malik Boykin's lab](https://www.boykinlab.com/) in Cognitive, Linguistic, and Psychological Sciences at Brown, we are studying people's preferred definition of fairness and what social and algorithmic factors influence these preferences. To power these tools, we are also developing techniques to interpolate between definitions of fairness. @@ -46,7 +46,9 @@ In collaboration with [Malik Boykin's lab](https://www.boykinlab.com/) in Cognit Task Level Fairness ^^^ -In this project, we examine how fairness can be evaluated at the task and problem level in order to develop heuristics for the feasibility of a fair model prior to training. +In this project, we examine how fairness can be evaluated at the task and problem level in order to develop heuristics for the feasibility of a fair model prior to training. + +This project will produce a Python library that anyone can use in the EDA step of their project. +++ @@ -61,7 +63,7 @@ In this project, we examine how fairness can be evaluated at the task and proble Model-Based Fairness Intervention Assessment ^^^ -In this project, we are using bias models to evalute the effectiveness of different types of fair machine learning interventions. +In this project, we are using bias models to evalute the effectiveness of different types of fair machine learning interventions. We hope to use the insights from this to provide data scientists with more actionable advice on how to select a fairness intervention. +++ @@ -75,11 +77,11 @@ In this project, we are using bias models to evalute the effectiveness of differ Wiggum ^^^ + Simpson's paradox inspired Fairness Forensics in collaboration with [the OU Data Lab](https://oudatalab.com/) +++ - [{fas}`house` ](https://fairnessforensics.github.io/wiggum) [{fab}`github` ](https://github.com/fairnessforensics/wiggum)