Many people think any given ML project is 99% training.
In reality, it’s 50% evaluation, 40% data cleaning, 8% integration, and 2% training.
The first two set the noise floor for learning. No ML magic matters; the model cannot lower the noise floor, as that’s the optimal bound of Shannon encoding of your data.
Thus, not a single day goes by without me thinking about ontology. Even the old labels have to be constantly reviewed.
Happy pi approximation day!
Today's date, 22 July, can also be written as the fraction 22/7, which is equal to 3.14285714, an approximation of π, correct to two decimal places.
Very happy to share our work out on MRI clearly showing the effects of distortion correction on ASL perfusion imaging using white paper consensus acquisition 👇
https://t.co/DMUG6zdmLv
Happy to share our recent paper published in MRM. 😄 @RitaGNunes@_nunoadasilva António Ferreira Anthony Price @laseeb_isr
Open-source myocardial T1 mapping with simultaneous multi-slice acceleration
#opensource#CMR#qMRI
Have a look: https://t.co/EDpW68ix1u
Did you know that Wilhelm Röntgen was unfairly expelled from school after being accused of producing a caricature of one of the teachers, which was in fact done by someone else.
He went onto discover X-rays and be awarded the very first #NobelPrize in Physics.
"Build your opinion on facts, data and knowledge.”
Chemistry laureate Ben Feringa urges us all to build our opinions on facts, data and knowledge.
Watch the full student roundtable here: https://t.co/W0v50bv06R
#NobelPrize
Who is more susceptible to disinformation? When? Why? If interested in our approach, you can now read this excellent short piece. Thank you @ERC_Research and Laura Bauer for such a clear highlight! @lip_pt@SpacLab
To introduce “IntelligentCare: Intelligent Multimorbidity Management System” we had @_nunoadasilva from Hospital da Luz Learning Health. The project is focused on developing a patient centric solution to help manage multimorbidity condition.
More: https://t.co/3wjJ6nMwBe
Robert FitzRoy (sim, o do Beagle) inventou a palavra "forecast". Queria estudar o clima para que barcos não fossem surpreendidos por tempestades.
Dados, análise, telégrafos, alerta
150 anos depois, porque não criar um "IPMA" para doenças infecciosas? 1/
https://t.co/lhFCgM4eby
I'm looking for a post-doctoral research fellow to join my team at MIITT/UM to work on an NIH-funded project in quantitative and rapid CMR... DM or email me if interested, and please RT :)