Who am I to say?

My Mission Statement

A year ago the Trump Administration traumatically separated thousands of children from their parents. I understand traumatic separation. Four years ago, I was suddenly separated from my six-year-old daughter and eight-year-old son.

Despite cases at the highest courts in Germany and the EU and the advice of all of Germany’s leading child’s rights lawyers, nothing changed. I haven’t heard my daughter’s voice in four years. By all reports, my son is traumatized.

German experts say this is not uncommon and not discussed. Research is rare, despite systemic child separations going back 80 years. I must change that.

I will develop the necessary interdisciplinary skills – starting with my current data science master’s at Harvard, followed by a child psychology master’s and a Ph.D. in computational social science (learning adult psychology, sociology, and advanced machine learning) at similar universities. Through this, I will develop independently-replicated, peer-reviewed quantitative research and data on the severity and pervasiveness of family separations – for my kids, and for what I estimate is millions of affected kids and families to start the serious policy changes to end it.


My name is Jeff Winchell and I live in Cambridge, MA. I am a dad, computational social scientist (formerly a data scientist), Harvard grad student, sporadic comedy writer, former racer (kind of an expensive hobby) and sometimes slam poet.

I work as a research assistant for five Harvard processors/researchers doing regression, NLP, RNN work in R, python, SQL Server 2019 and Stata to classify Japanese text level, quantify the semantic amount of environmental social and governance text in hedge fund documents, model the course of arguments in political conversations/videos, replicate studies on the effectiveness of international agreements and disambiguate the effects of age, period and cohort used in sociology studies. I am also working on a deep learning class project to classify emotions in videos.


You can reach me at Jeff_Winchell@g.Harvard.Edu, on Skype at live:CompSocialSci or even by an actual phone call or a meeting for coffee. How non-virtual. 😉

Returning to the theme of the above graphic, I’ve taken 5 courses towards my master’s – Big Data Analytics, Introduction to Data Science, Deep Learning, Advanced Scientific Computing and Probabilistic Programming. This semester I am taking Advanced Topics in Data Science and Advanced Quantitative Research Methodology (Political Science) and I’m auditing Machine Learning for Natural Language Processing. I had taken about 70 classes at Northwestern by the time I got my bachelor’s degree in math.

I’ve also completed 72 MOOCs over the past 6 years (and continue to take them concurrent with my master’s degree work). Here’s a list of my favorites (data science related courses listed first). In my view the first class in this list should be part of the required curriculum before taking any machine learning or deep learning class anywhere.

CaltechLearning from Data
Eindhoven Technische InstituteProcess Mining: Data Science in Action
DeepLearning.AIImproving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization
StanfordMachine Learning
DeepLearning.AINeural Networks and Deep Learning

Convolutional Neural Networks

MITIntroduction to Computer Science and Programming Using Python
ColumbiaBig Data in Education
MichiganSocial Network Analysis
Johns HopkinsR Programming
Univ of IL-Urbana-ChampaignText Mining and Analytics
Univ of Texas-AustinLinear Algebra – Foundations to Frontiers
HarvardIntroduction to Linear Models and Matrix Algebra
Univ of IL-Urbana-ChampaignPattern Discovery in Data Mining
Johns HopkinsPractical Machine Learning
Steve BlankHow to Build a Startup
UC-San DiegoLearning How to Learn: Powerful mental tools to help you master tough subjects
WesleyanSocial Psychology
Ohio StateCalculus Two: Sequences and Series
Case Western ReserveInspiring Leadership through Emotional Intelligence
StanfordHow to Learn Math: For Students
HarvardLeaders of Learning
Univ of LondonSupporting children with difficulties in reading and writing
StanfordCrash Course in Creativity
StanfordGame Theory
VirginiaEffective Classroom Interactions: Supporting Young Children’s Development
UC-BerkeleyThe Science of Happiness
HarvardUnlocking the Immunity to Change: A New Approach to Personal Improvement
Higher School of EconomicsIntroduction to Neuroeconomics: how the brain makes decisions
EdinburghThe Clinical Psychology of Children and Young People
Ohio StateCalculus One
Cal Institute of the ArtsIntroduction to Programming for Musicians and Digital Artists