I have been so concentrated on keeping my chill and my chakras aligned(TM) during this lockdown that I hadn't realized we get next Monday off and...

Hey #AI friends! Question for you: let's say I have a situation where I create a base model, then adapt that model to a local environment through unlabeled local data and user feedback. This isn't federated nor continuous learning. Do we have a name for this concept?

I’m definitely in the camp of defending rioting & looting as a legitimate, politically-informed response to state violence. I’ve tweeted on this topic many times & I’m always willing to engage people in good faith discussion of why I hold that position.

1) Premier François Legault decided not to sugar-coat the #pandemic on Tuesday, warning Quebecers there will be “a lot of uncertainty over the next two months.” Legault spoke of the #COVID19 resurgence in the regions. In this thread, I will focus on what he didn’t speak about.

I have something I need to get off my chest. If you work for an organization that supports LGBTQ+ people, you should be supporting LGBTQ+ journalists. Currently, there are ~2 orgs in the U.S. dedicated to recognizing the work of LGBTQ+ reporters. They have limited bandwidth.

I'm in a talk where a random yt dude just said "it's not political, it's about equity and diversity" and y'all clearly have not understood a single shit about #DEI, have you?

when you're dead inside but a colleague you really respect compliments your PR Review comments

"I have 13 PhDs" Do TV shows and movies understand how....getting a PhD works???

Who would've thought one of the most complex things in #coding is actually coming up with names that correctly abstract what your vars/funcs/classes represents? A colleague just shared with me what felt like an essay on why he named a class a certain way 🤖😂

Mercku Connectivity | WiFi & Smart Home and Aerial Partner to provide Best in Class Wi-Fi Sensing Services Un partenariat entre Mercku Connectivity | WiFi & Smart Home et Aerial pour offrir le meilleur service de détection Wi-Fi aerial.ai/resources/merc…

"The biggest gains are not from choosing the right model, but from framing a problem in the right way." towardsdatascience.com/my-advice-to-m…

Drop here seminal papers on the topic of comparing #MachineLearning and #DeepLearning models statistically. I'm already aware of Dietterich's (1998) "Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms", but I need mooooreeee 🧠🤖

I updated my #python distribution to the latest stable release and fell down... ⓓⓔⓟⓔⓝⓓⓔⓝⓒⓨ ⓗⓔⓛⓛ

re this: don't use stats testing. Use #ConfidenceIntervals--most metrics can be modeled using the normal approximation to a binomial distribution. When in industry, the models will be tested anyways by QA, so we can get away with being relatively lax in terms of statistics

The lab where I did my masters is looking for 2 postdocs and 1 research engineer to do research on music, social interactions, and neuroscience. I have nothing but good things to say about the MIMM institute and the people that run it ❤️ livelab.mcmaster.ca/jobs/