Director of Science Modeling at Climate LLC. Talking science, agriculture, games, and baseball. Opinions: my own. Mostly harmless.
St Louis, MO
Joined June 2008
- Tweets1.8K
- Following354
- Followers301
- Likes2.1K
Matt Dimmic retweeted
I like the idea of S-values or "Surprisal" values — I think it's a far more intuitive way for people to reason about evidence. I bet most people would be shocked that the standard criteria for statistical significance in science amounted to seeing about 4 heads in 4 flips.
After that Yepez assist from LF, just heard a literal As So Often Happens from Chip Caray, hope that makes the game story @dgoold
Sound journalism is so important in helping people understand the complexities of agriculture, and how their food gets made sustainably. Seems like a great opportunity.
Introducing our #Midwest Journalism Fund: We’re looking for 50 journalists interested in #environment & #agriculture across the Midwest to join our program.
You’ll gain access to training, mentorship and a $10,000 fund for new reporting. Apply by Sept 26: tinyurl.com/2y2txcx5
Batters should have the option to decline an intentional walk, or any four-pitch walk, until a strike is thrown.
Matt Dimmic retweeted
Fungicide trial analyzed with @fieldview region reports. Across 6 fields a 2.5bu gain vs checks. Done by simply selecting passes, but you can create any shape of region and get a breakdown by soil type, population, and more. Try it at bit.ly/TryClimateFV #fieldviewpartner
Matt Dimmic retweeted
Great summary of ML & DL, starting with perceptrons in 1950's.
And in particular, a great summary of the latest developments re architectures for SOTA computer vision:
CNNs -> visual transformers -> MLPs again without attention or convolutions -> CNNs again (ResNets tuned well)
Matt Dimmic retweeted
Calling all #AgTech startups! Win a chance to co-create with @Bayer4Crops through the @hellotmrc Innovation Challenge. Apply by August 22 to help shape the future of food: bit.ly/3wqwvux
🗓️SAVE THE DATE🗓️
After launching our Innovation Challenge with @Bayer4Crops, we're organising a webinar to present its topics, eligibility criteria & prizes: July 13th, 4pm CET!
A great opportunity to meet Bayer’s team & ask questions! Register for free👉bit.ly/2UBo2Y2
Check out this thread for some amazing machines at work. That sod harvester is so satisfying to watch.
Matt Dimmic retweeted
Calling on #AgTwitter to change up their seed game in #Plant21! Variable rate seeding combines the best of both worlds: #precisionag & #data.
Not convinced? Hear from two of our friends at @FieldView: bit.ly/2ZgWh6e
Teams may refuse any intentional or 4-pitch walk. Pitcher has to throw at least one strike.
This Post is from an account that no longer exists. Learn more
Matt Dimmic retweeted
I've said it before and I will say it again, there are some really cool opportunities in industry right now. Today's example: "Phenotypic Insights Lead" at Bayer Crop Science. jobs.bayer.com/job/Chesterfi…
Want to learn more about science careers outside of academia? Want to apply your genomics skills to crop improvement? This is a great summer opportunity with an outstanding team.
Attention PhD and Masters students: We are HIRING an intern for summer 2021 to help us dig deeper into our huge metagenome database. DM with questions! career.bayer.us/en/job/micro…
My opinion: the end state is few or no deaths each week.
We accomplish this by low population prevalence of the virus.
We can only get there by getting Rt < 1. Many other nations have shown us how to do this, in different ways. But we have to put in the work.
We will know we've failed when....<waves hands at surroundings>. Nationally we have 1,000+ deaths per day with Rt = 1 or higher. We don't seem to have a clear plan, or consistent metrics.