Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday,
March 2*, where Sharad Goel <https://5harad.com> (Harvard University)
presents "Designing Equitable Algorithms for Criminal Justice and Beyond."
*Abstract*
Machine learning algorithms are now used to automate routine tasks and to
guide high-stakes decisions, but, if not carefully designed, they can
exacerbate inequities. I’ll start by describing an evaluation of automated
speech recognition (ASR) tools, which power popular virtual assistants,
facilitate automated closed captioning, and enable digital dictation
platforms for health care. We find that five state-of-the-art ASR systems
-- developed by Amazon, Apple, Google, IBM, and Microsoft -- exhibited
substantial racial disparities, making twice as many errors for Black
speakers compared to white speakers, a gap we trace back to a lack of
diversity in the audio data used to train the models. I'll then describe
recent attempts to mathematically formalize fairness. I'll argue that some
of the most popular definitions, when used as a design principle, can,
perversely, harm the very groups they were created to protect. I'll
conclude by describing a general, consequentialist paradigm for designing
equitable algorithms that aims to mitigate the limitations of the dominant
approaches to building fair machine learning systems.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, March 2 at 12:10 - 1:30 pm.
(Bagged lunches available for pick-up at CGIS K354 *11:30 - 11:45 am*, for
the participants who responded to our previous survey. The CGIS cafe on the
first floor has been designated as an eating area, and participants may
also use outdoor spaces for lunch. Please be present at K354 by 12:10 pm
for the presentations.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday,
February 23*, where Soroush Saghafian
<https://scholar.harvard.edu/saghafian/home> (Harvard University) presents
"Ambiguous Dynamic Treatment Regimes: A Reinforcement Learning
Approach." The full paper is available here
<https://scholar.harvard.edu/saghafian/publications/ambiguous-dynamic-treatm…>
.
*Abstract*
A main research goal in various studies is to use an observational data set
and provide a new set of counterfactual guidelines that can yield causal
improvements. Dynamic Treatment Regimes (DTRs) are widely studied to
formalize this process and enable researchers to find guidelines that are
both personalized and dynamic. However, available methods in finding
optimal DTRs often rely on assumptions that are violated in real-world
applications (e.g., medical decision-making or public policy), especially
when (a) the existence of unobserved confounders cannot be ignored, and (b)
the unobserved confounders are time-varying (e.g., affected by previous
actions). When such assumptions are violated, one often faces ambiguity
regarding the underlying causal model that is needed to be assumed to
obtain an optimal DTR. This ambiguity is inevitable, since the dynamics of
unobserved confounders and their causal impact on the observed part of the
data cannot be understood from the observed data. Motivated by a case study
of finding superior treatment regimes for patients who underwent
transplantation in our partner hospital and faced a medical condition known
as New Onset Diabetes After Transplantation (NODAT), we propose a new
framework termed Ambiguous Dynamic Treatment Regimes (ADTRs), in which the
casual impact of treatment regimes is evaluated based on a “cloud” of
potential causal models. We then connect ADTRs to Ambiguous Partially
Observable Mark Decision Processes (APOMDPs) proposed by Saghafian (2018),
and consider unobserved confounders as latent variables but with ambiguous
dynamics and causal effects on observed variables. Using this connection,
we develop two Reinforcement Learning methods termed Direct Augmented
V-Learning (DAV-Learning) and Safe Augmented V-Learning (SAV-Learning),
which enable using the observed data to efficiently learn an optimal
treatment regime. We establish theoretical results for these learning
methods, including (weak) consistency and asymptotic normality. We further
evaluate the performance of these learning methods both in our case study
(using clinical data) and in simulation experiments (using synthetic data).
We find promising results for our proposed approaches, showing that they
perform well even compared to an imaginary oracle who knows both the true
causal model (of the data generating process) and the optimal regime under
that model.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, February 23 at 12:10 - 1:30 pm.
(Bagged lunches available for pick-up at CGIS K354 *11:30 - 11:45 am*, for
the participants who responded to our previous survey. The CGIS cafe on the
first floor has been designated as an eating area, and participants may
also use outdoor spaces for lunch. Please be present at K354 by 12:10 pm
for the presentations.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday,
February 16*, where Edward McFowland III
<https://www.hbs.edu/faculty/Pages/profile.aspx?facId=772797> (Harvard
University) presents "Anomalous Pattern Detection: A Novel Lens for
Scientific Inquiry."
*Abstract*
There has been a growing interest in the use of machine learning methods
for causal inference, which often involves adjusting or reappropriating
predictive models, with causality in mind. As an alternative, anomaly
detection methods offer a unique lens through which to conduct causal
inference, as the presence of a causal effect results in treatment group
units that appear anomalous in comparison to the control group. Moreover,
anomalous pattern detection intentionally localizes the presence of
treatment effects, which has tremendous value when the ultimate goal
involves hypothesis generation, understanding causal mechanisms, or
targeting subpopulations. As motivation, we will consider the
identification of subpopulations in randomized experiments with extremely
significant effects, and will consider other quasi-experimental settings as
time permits.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, February 16 at 12:10 - 1:30 pm.
(Bagged lunches available for pick-up at CGIS K354 *11:30 - 11:45 am*, for
the participants who responded to our previous survey. The CGIS cafe on the
first floor has been designated as an eating area, and participants may
also use outdoor spaces for lunch. Please be present at K354 by 12:10 pm
for the presentations.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn
Dear Applied Statistics Workshop Community,
Our next meeting of the semester will be at *12:10 pm (EST) Wednesday,
February 9*, where Tyler VanderWeele
<https://www.hsph.harvard.edu/tyler-vanderweele/> (Harvard University)
presents "The Global Flourishing Study - Seeking Analytic Input."
*Abstract*
The recently launched Global Flourishing Study
<https://hfh.fas.harvard.edu/files/pik/files/globalflourishingstudy_report.p…>
is
a longitudinal research study being carried out in collaboration between
scholars at the Human Flourishing Program <https://hfh.fas.harvard.edu/> at
Harvard's Institute for Quantitative Social Science, Baylor’s Institute for
Studies of Religion, Gallup, and the Center for Open Science.
The study will involve data collection for approximately 240,000
participants, from 22 geographically and culturally diverse countries, with
nationally representative samples within each country, and with annual data
collection on the same panel of individuals for five waves of data. The
survey includes a rich set of questions on well-being along with
demographic, social, economic, political, religious, personality,
childhood, community, health and character-based questions. The data will
constitute an open-access resource available to scholars throughout the
world. However, in addition to what are hoped to be diverse and
wide-ranging uses of the data, the primary research team intends to carry
out a series of coordinated parallel pre-registered analyses. The talk will
give an overview of the Global Flourishing Study itself and the flourishing
framework that motivated it, along with current analysis plans for the
coordinated pre-registered studies, with the aim of receiving critique,
suggestions, and feedback from the Applied Statistics Workshop
participants. Open questions will be put forward concerning appropriate
meta-analytic summaries, confounder control with a large number of highly
correlated indicators, and challenges of missing data and attrition, all
while respecting complex survey weights, the limitations of existing
software, and the desire to allow the utilization of multiple software
packages given the size and diversity of the primary research team.
*Where:* CGIS Knafel Building, Room K354
(See this link <https://map.harvard.edu/?bld=04471&level=9> for directions).
*When:* Wednesday, February 9 at 12:10 - 1:30 pm.
(Bagged lunches available for pick-up at CGIS K354 *11:30 - 11:45 am*, for
the participants who responded to our previous survey. The CGIS cafe on the
first floor has been designated as an eating area, and participants may
also use outdoor spaces for lunch. Please be present at K354 by 12:10 pm
for the presentations.)
*Zoom link*:
https://harvard.zoom.us/j/97004196610?pwd=eGFydkF5RDRjUlk5RVcyTjV6OStUQT09
(For the participants who cannot join the session physically.)
*Schedule of the workshop*:
https://projects.iq.harvard.edu/applied.stats.workshop-gov3009
Looking forward to seeing you all on Wednesday!
Best,
Sooahn