Welcome to the Flavours of Physics Challenge!\nWe expect this competition to be not very straightforward since we adapted metric from real physics analyses at LHC to bridge the gap between the Physics and the ML communities. Please read the challenge documentation (including extensive technical description) carefully. We hope you will find this competition enjoyable and rewarding, and in the end your joint efforts will help finding solutions that will open the door to the New Physics.
\n\nFeel free to contact us in case of trouble.
\n\nGood luck!
\n\nAndrey, Tatiana, Marcin, Marc-Olivier, Francesco, Thomas, Pavel, Evgeny, Alexey
", "messages": [ { "id": 86212, "postDate": "2015-07-20T23:03:38.847Z", "content": "Welcome to the Flavours of Physics Challenge!\nWe expect this competition to be not very straightforward since we adapted metric from real physics analyses at LHC to bridge the gap between the Physics and the ML communities. Please read the challenge documentation (including extensive technical description) carefully. We hope you will find this competition enjoyable and rewarding, and in the end your joint efforts will help finding solutions that will open the door to the New Physics.
\n\nFeel free to contact us in case of trouble.
\n\nGood luck!
\n\nAndrey, Tatiana, Marcin, Marc-Olivier, Francesco, Thomas, Pavel, Evgeny, Alexey
", "rawMarkdown": "Welcome to the **Flavours of Physics** Challenge!\r\nWe expect this competition to be not very straightforward since we adapted metric from real physics analyses at LHC to bridge the gap between the Physics and the ML communities. Please read the challenge documentation (including extensive [technical description][1]) carefully. We hope you will find this competition enjoyable and rewarding, and in the end your joint efforts will help finding solutions that will open the door to the New Physics.\r\n\r\nFeel free to contact us in case of trouble. \r\n\r\nGood luck!\r\n\r\nAndrey, Tatiana, Marcin, Marc-Olivier, Francesco, Thomas, Pavel, Evgeny, Alexey\r\n\r\n [1]: https://kaggle2.blob.core.windows.net/competitions/kaggle/4488/media/lhcb_description_official.pdf", "votes": 10 }, { "id": 86215, "postDate": "2015-07-20T23:52:26.527Z", "content": "Heh. It's going to be a busy summer!
", "rawMarkdown": "Heh. It's going to be a busy summer!", "votes": 5 }, { "id": 86862, "postDate": "2015-07-24T18:30:25.050Z", "content": "The organizers would like to remind everyone that they have added a new reference article that explores new ways to deal with unwanted correlation between classifier and a "spectator" (e.g. mass) variable.
\n\nLink is also available from the front page.
", "rawMarkdown": "The organizers would like to remind everyone that they have added a new reference [article][1] that explores new ways to deal with unwanted correlation between classifier and a \"spectator\" (e.g. mass) variable.\r\n\r\nLink is also available from the front page. \r\n\r\n [1]: http://iopscience.iop.org/1748-0221/10/03/T03002/pdf/1748-0221_10_03_T03002.pdf", "votes": 3 }, { "id": 86235, "postDate": "2015-07-21T05:24:22.910Z", "content": "This is gonna be a lot fun. A lot of goodies.
\n\nI smell xgb :) Anyone who would like to team?
", "rawMarkdown": "This is gonna be a lot fun. A lot of goodies.\r\n\r\n- chance your code actually going to be in LHC production \r\n- chance to meet the LHCb collaboration physicists\r\n- possible NIPS workshop submission!!!\r\n\r\nI smell xgb :) Anyone who would like to team?", "votes": 4 }, { "id": 95189, "postDate": "2015-10-05T19:22:10.057Z", "content": "Quick reminder that the first submissions & team merger deadline is in about 4.5 hours!
\n\nWe simulate SM up to the tau lepton production. Then the tau simulated to decay into 3 muons without any explicit BSM assumption.
\n\nAs you pointed our there are many BSM scenarios and simulating each of them is not really possible. What we do is to simulate so called flat phase space decay, this means that the tau decays just based on it's kinematics not properties not any actual physics model. Later on after we finish our analysis we can interpret this result in any BSM model just by re-weighting the phase space. We did this using effective field theory approach. If you are interested in this you can look in the attachment to this post.
\n\nIn the challenge you don't need to care about any lhcb software(like the VM you mentioned) or our simulation. We tried to digest as much of physics for you as possible so can have fun with machine learning.
\n\nLet me know if you need further explanation.
", "rawMarkdown": "We simulate SM up to the tau lepton production. Then the tau simulated to decay into 3 muons without any explicit BSM assumption. \r\n\r\nAs you pointed our there are many BSM scenarios and simulating each of them is not really possible. What we do is to simulate so called flat phase space decay, this means that the tau decays just based on it's kinematics not properties not any actual physics model. Later on after we finish our analysis we can interpret this result in any BSM model just by re-weighting the phase space. We did this using effective field theory approach. If you are interested in this you can look in the attachment to this post. \r\n\r\nIn the challenge you don't need to care about any lhcb software(like the VM you mentioned) or our simulation. We tried to digest as much of physics for you as possible so can have fun with machine learning.\r\n\r\nLet me know if you need further explanation.", "votes": 2 }, { "id": 87281, "postDate": "2015-07-28T09:00:22.687Z", "content": "I guess the easy part will be the prediction. Starting with the tools and entering the competition is the real challenge here (unless you click on a script indeed);=)
", "rawMarkdown": "I guess the easy part will be the prediction. Starting with the tools and entering the competition is the real challenge here (unless you click on a script indeed);=)" }, { "id": 86378, "postDate": "2015-07-21T20:19:10.347Z", "content": "[quote=Deep;86235]
\n\nThis is gonna be a lot fun. A lot of goodies.
\n\nI smell xgb :) Anyone who would like to team?
\n\n[/quote]
\n\nI would love to team up Deep. I am a newbie to machine learning though, but would love to learn and work with you :)
", "rawMarkdown": "[quote=Deep;86235]\r\n\r\nThis is gonna be a lot fun. A lot of goodies.\r\n\r\n- chance your code actually going to be in LHC production \r\n- chance to meet the LHCb collaboration physicists\r\n- possible NIPS workshop submission!!!\r\n\r\nI smell xgb :) Anyone who would like to team?\r\n\r\n[/quote]\r\n\r\n\r\nI would love to team up Deep. I am a newbie to machine learning though, but would love to learn and work with you :)" }, { "id": 86242, "postDate": "2015-07-21T06:57:16.430Z", "rawMarkdown": "", "votes": 2, "isDeleted": true } ], "comments": [ { "id": 86215, "author_name": "inversion", "author_url": "", "post_date": "2015-07-20T23:52:26.527000", "content": "Heh. It's going to be a busy summer!
", "votes": 5, "replies": [] }, { "id": 86862, "author_name": "Wendy Kan", "author_url": "", "post_date": "2015-07-24T18:30:25.050000", "content": "The organizers would like to remind everyone that they have added a new reference article that explores new ways to deal with unwanted correlation between classifier and a "spectator" (e.g. mass) variable.
\n\nLink is also available from the front page.
", "votes": 3, "replies": [] }, { "id": 86235, "author_name": "Deep", "author_url": "", "post_date": "2015-07-21T05:24:22.910000", "content": "This is gonna be a lot fun. A lot of goodies.
\n\nI smell xgb :) Anyone who would like to team?
", "votes": 4, "replies": [] }, { "id": 95189, "author_name": "Ana Lina Morelia", "author_url": "", "post_date": "2015-10-05T19:22:10.057000", "content": "Quick reminder that the first submissions & team merger deadline is in about 4.5 hours!
\n\nWe simulate SM up to the tau lepton production. Then the tau simulated to decay into 3 muons without any explicit BSM assumption.
\n\nAs you pointed our there are many BSM scenarios and simulating each of them is not really possible. What we do is to simulate so called flat phase space decay, this means that the tau decays just based on it's kinematics not properties not any actual physics model. Later on after we finish our analysis we can interpret this result in any BSM model just by re-weighting the phase space. We did this using effective field theory approach. If you are interested in this you can look in the attachment to this post.
\n\nIn the challenge you don't need to care about any lhcb software(like the VM you mentioned) or our simulation. We tried to digest as much of physics for you as possible so can have fun with machine learning.
\n\nLet me know if you need further explanation.
", "votes": 2, "replies": [] }, { "id": 87281, "author_name": "WhizWilde", "author_url": "", "post_date": "2015-07-28T09:00:22.687000", "content": "I guess the easy part will be the prediction. Starting with the tools and entering the competition is the real challenge here (unless you click on a script indeed);=)
", "votes": 0, "replies": [] }, { "id": 86378, "author_name": "krishnateja", "author_url": "", "post_date": "2015-07-21T20:19:10.347000", "content": "[quote=Deep;86235]
\n\nThis is gonna be a lot fun. A lot of goodies.
\n\nI smell xgb :) Anyone who would like to team?
\n\n[/quote]
\n\nI would love to team up Deep. I am a newbie to machine learning though, but would love to learn and work with you :)
", "votes": 0, "replies": [] }, { "id": 86242, "author_name": "", "author_url": "", "post_date": "2015-07-21T06:57:16.430000", "content": "", "votes": 2, "replies": [] } ], "raw_markdown_by_id": { "86212": "Welcome to the **Flavours of Physics** Challenge!\r\nWe expect this competition to be not very straightforward since we adapted metric from real physics analyses at LHC to bridge the gap between the Physics and the ML communities. Please read the challenge documentation (including extensive [technical description][1]) carefully. We hope you will find this competition enjoyable and rewarding, and in the end your joint efforts will help finding solutions that will open the door to the New Physics.\r\n\r\nFeel free to contact us in case of trouble. \r\n\r\nGood luck!\r\n\r\nAndrey, Tatiana, Marcin, Marc-Olivier, Francesco, Thomas, Pavel, Evgeny, Alexey\r\n\r\n [1]: https://kaggle2.blob.core.windows.net/competitions/kaggle/4488/media/lhcb_description_official.pdf", "86215": "Heh. It's going to be a busy summer!", "86862": "The organizers would like to remind everyone that they have added a new reference [article][1] that explores new ways to deal with unwanted correlation between classifier and a \"spectator\" (e.g. mass) variable.\r\n\r\nLink is also available from the front page. \r\n\r\n [1]: http://iopscience.iop.org/1748-0221/10/03/T03002/pdf/1748-0221_10_03_T03002.pdf", "86235": "This is gonna be a lot fun. A lot of goodies.\r\n\r\n- chance your code actually going to be in LHC production \r\n- chance to meet the LHCb collaboration physicists\r\n- possible NIPS workshop submission!!!\r\n\r\nI smell xgb :) Anyone who would like to team?", "95189": "Quick reminder that the first submissions & team merger deadline is in about 4.5 hours! \r\n\r\n- You need to make at least one submission before the deadline in order\r\n to compete in the remainder of this competition. \r\n- If you'd like to join forces with another data scientist or team of data scientists, you need to make it official before this time.", "86251": "We simulate SM up to the tau lepton production. Then the tau simulated to decay into 3 muons without any explicit BSM assumption. \r\n\r\nAs you pointed our there are many BSM scenarios and simulating each of them is not really possible. What we do is to simulate so called flat phase space decay, this means that the tau decays just based on it's kinematics not properties not any actual physics model. Later on after we finish our analysis we can interpret this result in any BSM model just by re-weighting the phase space. We did this using effective field theory approach. If you are interested in this you can look in the attachment to this post. \r\n\r\nIn the challenge you don't need to care about any lhcb software(like the VM you mentioned) or our simulation. We tried to digest as much of physics for you as possible so can have fun with machine learning.\r\n\r\nLet me know if you need further explanation.", "87281": "I guess the easy part will be the prediction. Starting with the tools and entering the competition is the real challenge here (unless you click on a script indeed);=)", "86378": "[quote=Deep;86235]\r\n\r\nThis is gonna be a lot fun. A lot of goodies.\r\n\r\n- chance your code actually going to be in LHC production \r\n- chance to meet the LHCb collaboration physicists\r\n- possible NIPS workshop submission!!!\r\n\r\nI smell xgb :) Anyone who would like to team?\r\n\r\n[/quote]\r\n\r\n\r\nI would love to team up Deep. I am a newbie to machine learning though, but would love to learn and work with you :)", "86242": "" } }