This channel features highlights of events and programming hosted by the Institute for Applied Computational Science (IACS), Harvard University's interdisciplinary graduate program and community for computational science and engineering. Website: iacs.seas.harvard.edu Facebook: facebook.com/HarvardIACS Twitter: @Harvard_IACS
One of a kind! This is just what I need: having listened to too many talks about how fancy the methods are, this is the one makes it clear about what do NOT work and why.
Amazing the capabilities of this, eg. imagine a world where the doc talks to a machine about the symptoms or health prevention goals and it creates medicinal capsules right for the patient, according to its medical logs
Very informative talk 💯, I really like the way she keeps questioning herself and push the research further. Many questions she put up and answered actually echo to my concerns, which makes me can smoothly understand this talk!
This is great, I've watched it two times, and am looking forward to any update! The factory farming industry cages and kills 60 billion animals each year, the amount of suffering is extreme, unprecedented and unbearable, really hope that you or anyone can end this as soon as possible!
So Karim means that Guiseppe is a recipe scouting bot? Taking a look at the patent it looks like that, and the actual products they have confirms that too (mostly copies of existing products) But I'm missing all that about how Notco's AI is looking at molecules of ingredients to come up with new ways of creating plant-based food, where is that?.
I'm pretty sure this is going to be one the biggest revolutions about food-tech! I'm very proud of being a Chilean computer engineer and to know that we have startups that are focusing in real world problems. Great mission, great technology and great group of people at NotCo, I hope one day I can work there and meet Giuseppe :D
on the Harvard Extension School's webpage, it seems like AM207 was already closed. Anybody know how can I watch the recorded video of this course? thx.
I'm a layperson though not completely ignorant of the scientific magic we now have since computers can learn. What I got from your presentation is ... this is just the tip of the iceberg. And the biggest challenge is structuring the data. Thank you.
the calibration of bnn's uncertainty appeared to be poor because model uncertainty (epistemic) is not all you need. you need to capture the heteroscedestic aleatoric uncertainty too.