Project status update // Actualización del estado del proyecto
Message boards :
News :
Project status update // Actualización del estado del proyecto
Message board moderation
Author | Message |
---|---|
Send message Joined: 18 Mar 15 Posts: 281 Credit: 2,748,608 RAC: 7 |
Dear volunteers, I am writing to update you on how the different experiments we had launched have been going and the next steps we will take. On the one hand, we have finished (at least temporarily) the optimization of our model. As we had already told you, with the results that you have sent us, we managed to greatly improve how our model represented different physiological markers, but a problem that we thought we had solved had reappeared. We have tried different options but it reappeared in all of them. That forces us to stop and take steps back in the definition of the model, so, for the moment, we will not be launching new simulations in NHuVe. On the other hand, as we had told you, we are trying to use our optimization algorithm in another existing model. The results seem to be very good, but we have to work in parallel on their validation. In the coming weeks we will be working on requesting funding from the Government of Aragon for a new research project that will use DENIS as a computing platform, so you may have noticed that our activity has decreased a little. If this goes well (hopefully!) we will keep you informed, but keep in mind that these things are slow so the answer will take months to arrive. Thank you all very much for staying there and for your support. Sincerely, Jesus. ================================================================== Estimados voluntarios: Os escribo para actualizar cómo han ido yendo los diferentes experimentos que teníamos lanzados y los próximos pasos que seguiremos. Por un lado, hemos terminado (al menos temporalmente) la optimización de nuestro modelo. Como ya os habíamos comentado, con los resultados que nos habéis ido mandando conseguíamos mejorar mucho cómo nuestro modelo representaba diferentes marcadores fisiológicos pero había reaparecido un problema que creíamos haber solucionado. Hemos probado diversas opciones pero en todas ellas reaparecía. Eso nos obliga a parar y dar pasos atrás en la definición del modelo, por lo que, por el momento, no estaremos lanzando nuevas simulaciones en NHuVe. Por otro lado, como os habíamos adelantado, estamos probando a utilizar nuestro algoritmo de optimización en otro modelo ya existente. Los resultados parecen ser muy buenos, pero tenemos que ir trabajando en paralelo en la validación de los mismos. En las próximas semanas estaremos trabajando en pedir financiación al Gobierno de Aragón para un nuevo proyecto de investigación que utilizará DENIS como plataforma de cómputo por lo que habréis notado que ha bajado un poco nuestra actividad. Si esto saliera bien (¡ojalá!) os mantendremos informados, pero contad que estas cosas son lentas así que la respuesta tardará meses en llegar. Muchas gracias a todos por seguir ahí y por vuestro apoyo. Atentamente, Jesús. Jesús Carro Universidad San Jorge @InSilicoHeart |
Send message Joined: 9 Apr 15 Posts: 171 Credit: 1,552,856 RAC: 1 |
The release, after some times, of the beta app is related to the problems you encountered?? Are you trying to create a new app? |
Send message Joined: 18 Mar 15 Posts: 281 Credit: 2,748,608 RAC: 7 |
No, it is the same app, but we are trying a parameters in the simulation far away from the usuall ones. I don't know if it could generate any problem and I prefer to send in the Beta. We are testing some modifications in the optimization algorithm, not in the model. It will requiere a lot of time. Best, Jesús. Jesús Carro Universidad San Jorge @InSilicoHeart |
Send message Joined: 9 Apr 15 Posts: 171 Credit: 1,552,856 RAC: 1 |
No, it is the same app, but we are trying a parameters in the simulation far away from the usuall ones. I don't know if it could generate any problem and I prefer to send in the Beta. Great idea!! We are testing some modifications in the optimization algorithm, not in the model. It will requiere a lot of time. No problem. Our cpus are ready to work for you!! |
Send message Joined: 25 Dec 23 Posts: 2 Credit: 1,828,491 RAC: 5 |
Thank you for the update. Very happy to have discovered this project and happy to crunch for it. |
Send message Joined: 5 Apr 23 Posts: 1 Credit: 137,758 RAC: 2 |
Thank you for the update, I am very excited to be crunching denis@home! |
Send message Joined: 23 Jan 24 Posts: 1 Credit: 32,878 RAC: 0 |
now we have started to run the performance of our servers on your projects, I hope you will sleep soon with the news :) |
Send message Joined: 28 Jun 15 Posts: 8 Credit: 574,508 RAC: 1 |
Greetings! I have received new WU. (01-Feb-2024) ( but I had removed DENIS projects from two other computers running BOINC after reading the news. The names of the new WU are ..._ORdInc_k_5-Drouin_MID_2000-conf_876.xml ...ORdInc_k_5-Pieske_500-conf_376.xml ...ORdInc_k_5-TESZT_Block_K1_75-conf_176.xml ...ORdInc_k_5-Schmidt_575-conf_576.xml ...ORdInc_k_5-CLVPET_Block_Kr_70-conf_677.xml . Any news about these? Can WU be highly produced now? THANKS in advance, Jay |
Send message Joined: 18 Mar 15 Posts: 281 Credit: 2,748,608 RAC: 7 |
Hi Jay, While we are preparing the new simulations, we are testing some modifications in the optimization algorithm. In particular, the ones related to ORdInc, are working quite well. The differences are related on how we modify the region in which we look for new combinations of parameters. In the original one, we reduce the zone if the approximation was bad, and we increase if a parameter was in the limit of the region. In the ORdInc, we are also increasing if the approximation is really good. The smallest error has been found with this new approximation, but this result needs to be validated out of the optimization... we need to be cautious By now, not, WU will not be highly produced yet. Best, Jesús. Jesús Carro Universidad San Jorge @InSilicoHeart |
Send message Joined: 28 Jun 15 Posts: 8 Credit: 574,508 RAC: 1 |
Jesus, Thank you for both the update and explanation!!! Jay |