Data Science Engineer @ IBM — 🏳️🌈 + 👨🏽💻 + 🥐 — all opinions are my own
Joined May 2020
- Tweets40
- Following44
- Followers26
- Likes28
Pinned Tweet
My response to the embarrassing opinion article "The Myth of Systemic Racism in Police" by @HMDatMI published in @WSJopinion @WSJ medium.com/@joeygibli/modern… #blacklivesmatter
call my speedy spice — if you’re doing a custom distance function literally just throw @numba_jit at it
thinking about the time my client said “kicking tires” to mean procrastinate, then five minutes later he said “kick tires” to mean get started
Random data science idea! Compare the similarity of two non-parametric models by using Kendall Tau metric on the top-N features
Excited to be a data science engineer on this project! @urbaninstitute and @IBM Data Science and AI Elite are collaborating to predict and characterize neighborhood changes across the USA using machine learning ibm.com/blogs/journey-to-ai/…
A fun #datascience idea: are you concerned that your model is biased? Try using #TSNE to embed your observations in two dimensions then color by whether they’re correct. If you can see clusters of correct/incorrect data points this may be evidence that your model is biased
Proud to announce that I’ll be serving as co-lead of the NYC LGBT+ Business Resource Group at IBM! Looking forward to building an accepting community :) #inclusiveIBM
Excited to be attending @OutandEqual Virtual Summit this year! Can’t wait to represent @IBM and engage with some of my idols in tech and media :) #inclusiveIBM
Just finished a 6-week engagement incorporating machine learning into a client’s advertisement bidding process! Looking forward to the next project with #IBMDSE @IBMDataScience @IBMData
I don’t know who needs to hear this, but if you’re getting a lot of NaN when merging two dataframes try explicitly declaring types 🙃 #Python
Spent way too long debugging this, but type inference in Pandas is funky and you might end up with “mixed type” when reading from a CSV instead of “string” depending on the order of rows? @pandas_dev #Python
The United States certainly *feels* more divided than ever, wondering if there is a way to gauge the polarization of countries without polling...
