Fantástico video. Gracias. Por favor, ¿pudo pedirle al modelo la predicción de nuevos valores? Por ejemplo, para t=41,42,...,50. ¿Puedo suponer que el método funciona para una sola variable v en lugar de dos variables u1 y u2? Un buen reto práctico podría ser modelar el índice del fenómeno de El Niño, ONI INDEX de la NOAA y predecir los próximos 24 o 36 valores mensuales próximos. O, el anterior índice junto con el SOI, con el que está estrechamente correlacionado. ¿Tiene esto sentido? Gracias de nuevo.
Excelentes preguntas! Desafortunadamente, no podré responderlas en este momento ya que todavía estoy aprendiendo a usar estas herramientas. Me pondré en contacto con usted después de haber tenido algo de tiempo para conocer mejor estas herramientas.
I got my PhD in Physics about 40 years ago. I almost don't recognize the analysis techniques of today, that's how far they've advanced. But hey, at least I'll have something to keep my mind sharp in retirement!
Hi, thank you for the amazing content. I still in the way to find where to put my foot in science field and I found your video about scientific programming. It's really good to know about a new thing but I prefer to see it in literal form. Is there any paper or keyword to dig into this field?
Good question. To be honest, I haven't found any good resources. The term "scientific programming" , or "scientific computing", is not well defined since it is relatively new and it is evolving rapidly. In my experience, most of the literature available online is outdated.
then you take your newly learned NN home, happy it has now learned from the data. You start to play around with it and then you let the time progress beyond the set of data that was initially used, the model quickly separates of the reality. So for time evolving systems like physical system, this is of no help. at least no more than linear regresion or multiobjective optimization
![[08x13] Real-Time Fluid Dynamics Simulation in Julia using Waterlily.jl, GLMakie.jl and VS Code](http://i.ytimg.com/vi/hYbG7dFDCdU/mqdefault.jpg)
![[08x13] Real-Time Fluid Dynamics Simulation in Julia using Waterlily.jl, GLMakie.jl and VS Code](/img/tr.png)
![[08x10] Intro to Probabilistic Programming in Julia using Turing.jl and Pluto](http://i.ytimg.com/vi/ixHKOz2vnoA/mqdefault.jpg)
actually, you have more updated julia contents than julia channel itself.
Thanks
Thank you very much!
that's really amazing how you can determine the function and the input merely based upon the output.
All your videos are good enough
Fantástico video. Gracias. Por favor, ¿pudo pedirle al modelo la predicción de nuevos valores? Por ejemplo, para t=41,42,...,50. ¿Puedo suponer que el método funciona para una sola variable v en lugar de dos variables u1 y u2? Un buen reto práctico podría ser modelar el índice del fenómeno de El Niño, ONI INDEX de la NOAA y predecir los próximos 24 o 36 valores mensuales próximos. O, el anterior índice junto con el SOI, con el que está estrechamente correlacionado. ¿Tiene esto sentido? Gracias de nuevo.
Excelentes preguntas! Desafortunadamente, no podré responderlas en este momento ya que todavía estoy aprendiendo a usar estas herramientas. Me pondré en contacto con usted después de haber tenido algo de tiempo para conocer mejor estas herramientas.
I got my PhD in Physics about 40 years ago. I almost don't recognize the analysis techniques of today, that's how far they've advanced. But hey, at least I'll have something to keep my mind sharp in retirement!
All of this new stuff is pretty exciting! Sounds like a great hobby for retirement!
Hi, thank you for the amazing content. I still in the way to find where to put my foot in science field and I found your video about scientific programming. It's really good to know about a new thing but I prefer to see it in literal form. Is there any paper or keyword to dig into this field?
Good question. To be honest, I haven't found any good resources. The term "scientific programming" , or "scientific computing", is not well defined since it is relatively new and it is evolving rapidly. In my experience, most of the literature available online is outdated.
then you take your newly learned NN home, happy it has now learned from the data. You start to play around with it and then you let the time progress beyond the set of data that was initially used, the model quickly separates of the reality. So for time evolving systems like physical system, this is of no help. at least no more than linear regresion or multiobjective optimization
Thanks
Thank you!