DataDoses @Datadoses

Data science, ML, AI. Talk to me about the intersection of data and cyber security. Endlessly interested. Joined September 2021

Tweets

259
Followers

318
Following

620
Likes

4K

DataDoses @Datadoses

a year ago

Fantastic panel of women in data and AI at the databricks summit! #databricks #dais2024

0 1 1 328 0

Download Image

DataDoses @Datadoses

2 years ago

Not data inputs, not data outputs, but a secret third thing. Learn embeddings here.

vicki @vboykis

2 years ago

Not data inputs, not data outputs, but a secret third thing. Learn embeddings here.

51 456 2K 424K 2K

0 0 1 157 0

DataDoses @Datadoses

3 years ago

Want to help out those who may have gone through layoffs or are otherwise struggling, but not sure how? Consider this toy drive!

Adam Karpiak @Adam_Karpiak

3 years ago

Want to help out those who may have gone through layoffs or are otherwise struggling, but not sure how? Consider this toy drive!

21 70 210 0 2

1 0 1 0 0

DataDoses @Datadoses

3 years ago

To negotiate the job offer or not to negotiate the job offer... I wish I was as confident as those who insist you should always negotiate.

0 0 2 0 0

DataDoses @Datadoses

3 years ago

Full stack of data science goodness here.

8 51 393 0 169

Download Image

To build a metrics framework, ask 4 questions: ➜ What is this function’s output? ➜ What inputs does this function need to create that output? ➜ What steps does this function follow to turn those inputs to the output? ➜ What systems are involved in those steps?

2 3 9 0 7

Tom Carpenter, PhD (@tcarpenter.bsky.social) @tcarpenter216

3 years ago

In case you wondered why the “database icon” looks like a cylinder

Ben Larson, Ph.D. @blarson424

3 years ago

In case you wondered why the “database icon” looks like a cylinder

2 5 31 0 2

Download Image

2 73 270 0 15

DataDoses @Datadoses

3 years ago

"Initially remote" is not a draw at all.

0 0 1 0 0

DataDoses @Datadoses

3 years ago

When in the mid-career job exploration phase, am I better off spending time on projects or on hackerank type sites for coding interview practice? (With a job already, assume i only have time for one or the other, and even that is limited)

1 1 2 0 2

DataDoses @Datadoses

3 years ago

Casually looking at new roles, but turns out being ghosted by a recruiter for one role and then having my application be rated as "not meeting qualifications" for another still stings. 🫤

0 0 2 0 1

DataDoses @Datadoses

3 years ago

It still catches me off guard how often cleaning/learning the data as prep for building a model leads to insights that render the need for a model obsolete.

0 0 3 0 0

DataDoses @Datadoses

3 years ago

Found data leakage in a suspiciously high-performing model we were prepping to deploy. This means I've graduated to senior data scientist, right?

0 0 5 0 2

DataDoses @Datadoses

3 years ago

Read a LinkedIn post from a DS manager pontificating on the differences in candidates who have a data science degree vs a degree in software engineering and which he considered more "qualified." So I checked his profile. His degree, 30 years previous? Industrial psychology.

2 1 16 0 0

DataDoses @Datadoses

3 years ago

In addition to their "casually looking for new roles, recruiters can see" setting on @LinkedIn, I wish there was a place to list min salary you'd need to consider leaving current role. A lot of recruiters might ignore it, but it could save time/effort on both sides.

0 0 3 0 0

DataDoses @Datadoses

4 years ago

Multiple range joins on time series data kicked my butt this week.

0 0 2 0 0

DataDoses @Datadoses

4 years ago

#100DaysOfCode, #66daysofdata - these initiatives are great, and i am all for anything that helps you stay consistent and engaged. But keep in mind - while establishing a habit is helpful, the outcome is what matters. What did you learn? What are you building?

1 2 20 0 1

DataDoses @Datadoses

4 years ago

My daughter used code.org to build some games over the holidays. Every time she added something, she said it only worked if she commented her code first. I did not correct her.

1 0 6 0 0

DataDoses @Datadoses

4 years ago

I don't come from a software engineering background, so while I knew incorporating tests was important in data science, I haven't known where to start. Big thanks to @pmbaumgartner for his clear examples - will be trying these out in my code!

1 0 5 0 0

DataDoses @Datadoses

4 years ago

I've slowly been working to win over a director who has been dismissive of our data efforts in the past. This morning he reached out to ask for a follow up analysis on something I presented last week. Sometimes the data science wins don't come from the data at all.

0 0 4 0 0

DataDoses @Datadoses

4 years ago

I built a streamlit app that has a numbered list of steps and for #8, I put "I forget what 8 was for" and my post about the app has 12k+ views and dozens of replies and no one has mentioned #8. Just devastating.