Lightnews — Scholar-powered news

Tony Dunsworth, PhD

@trdunsworth.bsky.social

3.4K followers 5.4K following 270 posts

Data scientist practicing in a 911 centre. Hockey, cricket, and rugby enthusiast. Time series forecasting and bringing data science to PSAPs is my passion. NENA volunteer. Please no DMs unless you know me IRL!!

Posts Media Videos Starter Packs

Tony Dunsworth, PhD @trdunsworth.bsky.social · 3d

I can do that later today

Tony Dunsworth, PhD @trdunsworth.bsky.social · 3d

Were I still in my PhD program, this would be interesting.

1 2

Tony Dunsworth, PhD @trdunsworth.bsky.social · 4d

I've bought them, started before I found out about his community attitude. I can share a Google Drive folder with you. I'm reviewing to see how accurate he is. I'm being very very critical.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 7d

You could pick any 4 actress

Tony Dunsworth, PhD @trdunsworth.bsky.social · 7d

My grandfather fought fascists. I wouldn't be able to look him in the eye if I didn't.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 9d

If I weren't already following you, is follow you for that.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 9d

D appears to be a the correct response

1 4

Tony Dunsworth, PhD @trdunsworth.bsky.social · 9d

This sounds more scientifically valid than anything else I've heard.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 9d

I'm sorry to hear that. I hope y'all are ok afterwards.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 12d

I can't imagine using that phrase in an academic paper, so I'm glad it's being called out.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 15d

The folks needed to fix the flying cars were ain't got.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 15d

If I were in the area, I'd gladly join. I'm working on a quarto document using R right now.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 16d

I'm interested in reading the papers as they come out.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 16d

I'm unable to answer that question for fear of reprisals

1 2

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

If you're interested in the website, it's drddatascience.com My blog is linked in there as well. I never thought to do a tutorial as to how I created it.

Dr. D’s Data Science

drddatascience.com

1 1

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

If it were me, I'd likely impute, then subset so I have the data needed.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

The other idea I've been kicking around is updating my blog: drddatacience.com/blog talking about my dissertation. Many friends in the 911 community are interested, but the language is too academic and my data friends don't know as much about 911 centers and their data. Thoughts?

drddatacience.com

1 5

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

I would haunt anyone foolish enough to think of that for me.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

I'd seen that post. I have a few issues with Valeriy and his blanket statements and his insistence that only Conformal Prediction is the right answer. I've thought about going through his books carefully to see what else lies in wait.

1 2

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

I hadn't seen that article. However, you delicately addressed my thoughts of him being "difficult" much better than I could. He had a post on LinkedIn over the weekend that declared that an arXiv preprint describing a new model would replace ARIMA. I said and believe that's too early to suggest.

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

I use it as a daily driver for my blog and my report building. Absolutely love it!

1 1

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

What do you do when you MUST forecast something with data that isn't amenable to forecasting? I work with really challenging data, so it's in the back of my mind when I'm reading books or papers. I think some of his opinion pieces are meant to stir the pot, so I'm playing along with my data.

1 1

Tony Dunsworth, PhD @trdunsworth.bsky.social · 17d

I've read the first two chapters and it's not a bad read. If you follow him on LinkedIn or Medium, you'll note that he's convinced that anything other than Confirmation Prediction isn't good enough, but when he started discussing forecastability in chapter 2, I had some ideas....

1 2

Tony Dunsworth, PhD @trdunsworth.bsky.social · 18d

Many ideas this weekend, should I propose a book idea covering Quarto and its many uses? Start up a project covering Manokhin's book, valeman.gumroad.com/l/MasteringM... and translate some of the code into R where I can? Thoughts from my data friends?

🚀 Pro Edition → Book + Bonus Pack Mastering Modern Time Series Forecasting 🔥 The Complete Guide to Statistical, Machine Learning & Deep Learning Models in Python — Now with Premium Tools, Templates & ...

🔥 Pro Edition: Mastering Modern Time Series Forecasting (Early Access) The elite version of the book — trusted by data science leaders in 100+ countries.Unlock the premium toolkit behind today’s most ...

valeman.gumroad.com

2 3

Tony Dunsworth, PhD @trdunsworth.bsky.social · 20d

Been reading a new book on time series forecasting and I think I'm going to follow along with many of the techniques the author demonstrates because I've worked on some challenging data that I question how forecastable it is. Will make some posts about it in future