Hey all,
\nThis was a great learning experience, I'd like to thank
See you in the next one! Cheers!
", "messages": [ { "id": "1043262", "postDate": "10/08/2020 20:07:09", "content": "Hey all,
\nThis was a great learning experience, I'd like to thank
See you in the next one! Cheers!
", "rawMarkdown": "Hey all,\nThis was a great learning experience, I'd like to thank \n1. All the participants for the great discussions we had over the forum and within the notebooks\n2. OSIC team for organizing this great & purposeful competition\n3. Kaggle team for providing the environment where (1) and (2) got together :)\n\nSee you in the next one! Cheers!", "votes": null }, { "id": "1043538", "postDate": "10/09/2020 04:01:32", "content": "Personally, I think this is more of a recommendation question. I was very pleasantly surprised when I read this article of yours, thanks for your professional analysis, I really enjoyed it, respect.
", "rawMarkdown": "Personally, I think this is more of a recommendation question. I was very pleasantly surprised when I read [this article](https://www.kaggle.com/carlossouza/bayesian-experiments) of yours, thanks for your professional analysis, I really enjoyed it, respect.", "votes": null }, { "id": "1044012", "postDate": "10/09/2020 12:25:15", "content": "Many thanks, Carlos for your contributions throughout the competition, I have been following your posts and notebooks. There are many lessons learned in this competition for us and hopefully for participants as well. Looking forward to seeing you and your team in future competitions!
", "rawMarkdown": "Many thanks, Carlos for your contributions throughout the competition, I have been following your posts and notebooks. There are many lessons learned in this competition for us and hopefully for participants as well. Looking forward to seeing you and your team in future competitions!", "votes": null }, { "id": "1048993", "postDate": "10/14/2020 02:54:22", "content": "I learned a lot from your deep dive in Bayesian models and even signed up for a Russian bayesian statistics course :-). Full disclosure, I only got about 13 lectures in then had to drop haha but your notebooks were extremely informative, and I imagine myself returning to them.
", "rawMarkdown": "I learned a lot from your deep dive in Bayesian models and even signed up for a Russian bayesian statistics course :-). Full disclosure, I only got about 13 lectures in then had to drop haha but your notebooks were extremely informative, and I imagine myself returning to them.", "votes": null }, { "id": "1049036", "postDate": "10/14/2020 03:27:25", "content": "Thanks @wanping7 !
", "rawMarkdown": "Thanks @wanping7 !", "votes": null }, { "id": "1049039", "postDate": "10/14/2020 03:36:30", "content": "Thanks @authman !! Yeah, I dropped that course as well, too hard to follow :)
\nMy Probabilistic Machine Learning / Bayesian methods learning curve was significantly helped by these books:
I finally settled to Pyro as my choice for (Deep Universal) Probabilistic Programming language (over TF Probability or PyMC3), they have nice tutorials that also really helped! (Full disclosure, I became a contributor to the project :))
\nCheers!
", "rawMarkdown": "Thanks @authman !! Yeah, I dropped that course as well, too hard to follow :)\nMy Probabilistic Machine Learning / Bayesian methods learning curve was significantly helped by these books:\n- [Machine Learning: A Probabilistic Perspective](https://doc.lagout.org/science/Artificial%20Intelligence/Machine%20learning/Machine%20Learning_%20A%20Probabilistic%20Perspective%20%5BMurphy%202012-08-24%5D.pdf)\n- [Automating Inference, Learning, and Design using Probabilistic Programming](https://www.robots.ox.ac.uk/~twgr/assets/pdf/rainforth2017thesis.pdf)\n- [Bayesian Methods for Hackers](https://github.com/CamDavidsonPilon/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers)\n- [Probabilistic Models of Cognition](http://probmods.org/)\n\nI finally settled to [Pyro](http://pyro.ai/) as my choice for (Deep Universal) Probabilistic Programming language (over TF Probability or PyMC3), they have nice tutorials that also really helped! (Full disclosure, I became a contributor to the project :))\n\nCheers!", "votes": null }, { "id": "1049341", "postDate": "10/14/2020 10:49:47", "content": "Just wanted to let you know that there's someone out there, inspired from your ideas, used and built a Bayesian probabilistic model in TF factoring in epistemic and aleatoric uncertainties. The model scored 7.0 by predicting normal distribution over 133 weeks WITHOUT using the base FVC.. It was fun. Thanks for your contribution!
", "rawMarkdown": "Just wanted to let you know that there's someone out there, inspired from your ideas, used and built a Bayesian probabilistic model in TF factoring in epistemic and aleatoric uncertainties. The model scored 7.0 by predicting normal distribution over 133 weeks WITHOUT using the base FVC.. It was fun. Thanks for your contribution!", "votes": null }, { "id": "1049531", "postDate": "10/14/2020 14:04:31", "content": "Textbook for the new course I'll be taking..
", "rawMarkdown": "Textbook for [the new course](https://twitter.com/matvil/status/1315907788841811969) I'll be taking..", "votes": null } ], "comments": [ { "id": 1043538, "author_name": "wanping7", "author_url": "", "post_date": "10/09/2020 04:01:32", "content": "Personally, I think this is more of a recommendation question. I was very pleasantly surprised when I read this article of yours, thanks for your professional analysis, I really enjoyed it, respect.
", "votes": null, "replies": [ { "id": 1049036, "author_name": "carlossouza", "author_url": "", "post_date": "10/14/2020 03:27:25", "content": "Thanks @wanping7 !
", "votes": null, "replies": [] } ] }, { "id": 1044012, "author_name": "ahmedhshahin", "author_url": "", "post_date": "10/09/2020 12:25:15", "content": "Many thanks, Carlos for your contributions throughout the competition, I have been following your posts and notebooks. There are many lessons learned in this competition for us and hopefully for participants as well. Looking forward to seeing you and your team in future competitions!
", "votes": null, "replies": [] }, { "id": 1048993, "author_name": "authman", "author_url": "", "post_date": "10/14/2020 02:54:22", "content": "I learned a lot from your deep dive in Bayesian models and even signed up for a Russian bayesian statistics course :-). Full disclosure, I only got about 13 lectures in then had to drop haha but your notebooks were extremely informative, and I imagine myself returning to them.
", "votes": null, "replies": [ { "id": 1049039, "author_name": "carlossouza", "author_url": "", "post_date": "10/14/2020 03:36:30", "content": "Thanks @authman !! Yeah, I dropped that course as well, too hard to follow :)
\nMy Probabilistic Machine Learning / Bayesian methods learning curve was significantly helped by these books:
I finally settled to Pyro as my choice for (Deep Universal) Probabilistic Programming language (over TF Probability or PyMC3), they have nice tutorials that also really helped! (Full disclosure, I became a contributor to the project :))
\nCheers!
", "votes": null, "replies": [] }, { "id": 1049531, "author_name": "authman", "author_url": "", "post_date": "10/14/2020 14:04:31", "content": "Textbook for the new course I'll be taking..
", "votes": null, "replies": [] } ] }, { "id": 1049341, "author_name": "ramanareddyeceb", "author_url": "", "post_date": "10/14/2020 10:49:47", "content": "Just wanted to let you know that there's someone out there, inspired from your ideas, used and built a Bayesian probabilistic model in TF factoring in epistemic and aleatoric uncertainties. The model scored 7.0 by predicting normal distribution over 133 weeks WITHOUT using the base FVC.. It was fun. Thanks for your contribution!
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "1043262": "Hey all,\nThis was a great learning experience, I'd like to thank \n1. All the participants for the great discussions we had over the forum and within the notebooks\n2. OSIC team for organizing this great & purposeful competition\n3. Kaggle team for providing the environment where (1) and (2) got together :)\n\nSee you in the next one! Cheers!", "1043538": "Personally, I think this is more of a recommendation question. I was very pleasantly surprised when I read [this article](https://www.kaggle.com/carlossouza/bayesian-experiments) of yours, thanks for your professional analysis, I really enjoyed it, respect.", "1044012": "Many thanks, Carlos for your contributions throughout the competition, I have been following your posts and notebooks. There are many lessons learned in this competition for us and hopefully for participants as well. Looking forward to seeing you and your team in future competitions!", "1048993": "I learned a lot from your deep dive in Bayesian models and even signed up for a Russian bayesian statistics course :-). Full disclosure, I only got about 13 lectures in then had to drop haha but your notebooks were extremely informative, and I imagine myself returning to them.", "1049036": "Thanks @wanping7 !", "1049039": "Thanks @authman !! Yeah, I dropped that course as well, too hard to follow :)\nMy Probabilistic Machine Learning / Bayesian methods learning curve was significantly helped by these books:\n- [Machine Learning: A Probabilistic Perspective](https://doc.lagout.org/science/Artificial%20Intelligence/Machine%20learning/Machine%20Learning_%20A%20Probabilistic%20Perspective%20%5BMurphy%202012-08-24%5D.pdf)\n- [Automating Inference, Learning, and Design using Probabilistic Programming](https://www.robots.ox.ac.uk/~twgr/assets/pdf/rainforth2017thesis.pdf)\n- [Bayesian Methods for Hackers](https://github.com/CamDavidsonPilon/Probabilistic-Programming-and-Bayesian-Methods-for-Hackers)\n- [Probabilistic Models of Cognition](http://probmods.org/)\n\nI finally settled to [Pyro](http://pyro.ai/) as my choice for (Deep Universal) Probabilistic Programming language (over TF Probability or PyMC3), they have nice tutorials that also really helped! (Full disclosure, I became a contributor to the project :))\n\nCheers!", "1049341": "Just wanted to let you know that there's someone out there, inspired from your ideas, used and built a Bayesian probabilistic model in TF factoring in epistemic and aleatoric uncertainties. The model scored 7.0 by predicting normal distribution over 133 weeks WITHOUT using the base FVC.. It was fun. Thanks for your contribution!", "1049531": "Textbook for [the new course](https://twitter.com/matvil/status/1315907788841811969) I'll be taking.." }, "source": "meta" }