Here is a document detailing the complex math of one of the models given in https://www.kaggle.com/scirpus/handle-missing-values-not.\nHope you like it.\nn.b. The overlines are there to remind one that these are complex numbers!
", "messages": [ { "id": "740634", "postDate": "02/09/2020 15:46:22", "content": "Here is a document detailing the complex math of one of the models given in https://www.kaggle.com/scirpus/handle-missing-values-not.\nHope you like it.\nn.b. The overlines are there to remind one that these are complex numbers!
", "rawMarkdown": "Here is a document detailing the complex math of one of the models given in https://www.kaggle.com/scirpus/handle-missing-values-not.\nHope you like it.\nn.b. The overlines are there to remind one that these are complex numbers!", "votes": null }, { "id": "740753", "postDate": "02/09/2020 18:35:37", "content": "So edgy!
", "rawMarkdown": "So edgy!", "votes": null }, { "id": "740771", "postDate": "02/09/2020 19:00:34", "content": "More like so bored waiting for a non deep learning competition.\nI do Kaggle as a bit of fun rather than wanting to download an exabyte of image data ;)
", "rawMarkdown": "More like so bored waiting for a non deep learning competition.\nI do Kaggle as a bit of fun rather than wanting to download an exabyte of image data ;)", "votes": null }, { "id": "741422", "postDate": "02/10/2020 16:02:54", "content": "I pictured you writing that with blood on the walls of a prison cell. I'll take the bait: is there a logic behind it? Like algebraic stacks or qabalah or something.
", "rawMarkdown": "I pictured you writing that with blood on the walls of a prison cell. I'll take the bait: is there a logic behind it? Like algebraic stacks or qabalah or something.", "votes": null }, { "id": "741430", "postDate": "02/10/2020 16:12:20", "content": "It is a tool I wrote that you give it a target and then it uses darwinian evolution to get to a solution - this took over 2000 generations. It generates latex, r, c++ and python.
", "rawMarkdown": "It is a tool I wrote that you give it a target and then it uses darwinian evolution to get to a solution - this took over 2000 generations. It generates latex, r, c++ and python.", "votes": null }, { "id": "743111", "postDate": "02/11/2020 19:49:26", "content": "\n\nI really appreciate your contributions but, I still don’t understand Genetic.
\n\nI did ask you sometime ago some documentation, paper and I still would appreciate.
\n\nYour scripts are so obscure but efficient ;-)
\n\nRegards
", "rawMarkdown": "scirpus \n\nI really appreciate your contributions but, I still don’t understand Genetic.\n\nI did ask you sometime ago some documentation, paper and I still would appreciate.\n\nYour scripts are so obscure but efficient ;-)\n\nRegards", "votes": null }, { "id": "743115", "postDate": "02/11/2020 19:53:09", "content": "\n\n« this took over 2000 generations »
\n\nMister Darwin I presume ...
\n\nI would like to understand your model(s), really !
", "rawMarkdown": "scirpus \n\n« this took over 2000 generations » \n\nMister Darwin I presume ...\n\nI would like to understand your model(s), really !", "votes": null }, { "id": "744214", "postDate": "02/12/2020 16:57:50", "content": "Hey there - I did give you a pdf that expresses the code as pure math ;)
", "rawMarkdown": "Hey there - I did give you a pdf that expresses the code as pure math ;)", "votes": null }, { "id": "744353", "postDate": "02/12/2020 19:01:08", "content": "Really ?
\n\nSomething a « standard »human could understand ?
\n\nIt is anyway fascinating :-)
", "rawMarkdown": "Really ?\n\nSomething a « standard »human could understand ?\n\nIt is anyway fascinating :-)", "votes": null }, { "id": "746267", "postDate": "02/14/2020 19:51:35", "content": "I like the idea of embedding the \"missing\" values on the purely imaginary axis. Have you ever played around with sending them to different values
\n\n$$\n( 0+b_k \\cdot i ) \n$$
\n\nfor different features
\n\n$$\nX_k \\textrm{?}\n$$
", "rawMarkdown": "I like the idea of embedding the \"missing\" values on the purely imaginary axis. Have you ever played around with sending them to different values \n\n$$\n( 0+b_k \\cdot i ) \n$$ \n\nfor different features \n\n$$\nX_k \\textrm{?}\n$$", "votes": null }, { "id": "746273", "postDate": "02/14/2020 19:54:40", "content": "I haven't yet - the code is pretty new - I'll keep you posted!
", "rawMarkdown": "I haven't yet - the code is pretty new - I'll keep you posted!", "votes": null }, { "id": "746278", "postDate": "02/14/2020 19:56:12", "content": "\n\nI wanna be among them too...
", "rawMarkdown": "scirpus \n\nI wanna be among them too...", "votes": null }, { "id": "746290", "postDate": "02/14/2020 20:08:44", "content": "Awesome! Much appreciated.
\n\nOne further question: Without giving away your secret sauce --are you using a memetic algorithm per chance?
", "rawMarkdown": "Awesome! Much appreciated.\n\nOne further question: Without giving away your secret sauce --are you using a memetic algorithm per chance?", "votes": null }, { "id": "746314", "postDate": "02/14/2020 20:32:23", "content": "Straight Genetic Programming - Look at anything by John Koza - who also invented the lottery scratchcard!
", "rawMarkdown": "Straight Genetic Programming - Look at anything by John Koza - who also invented the lottery scratchcard!", "votes": null }, { "id": "746319", "postDate": "02/14/2020 20:38:38", "content": "Gotcha. Thanks for the reference!
", "rawMarkdown": "Gotcha. Thanks for the reference!", "votes": null } ], "comments": [ { "id": 740753, "author_name": "davidbnn92", "author_url": "", "post_date": "02/09/2020 18:35:37", "content": "So edgy!
", "votes": null, "replies": [ { "id": 740771, "author_name": "scirpus", "author_url": "", "post_date": "02/09/2020 19:00:34", "content": "More like so bored waiting for a non deep learning competition.\nI do Kaggle as a bit of fun rather than wanting to download an exabyte of image data ;)
", "votes": null, "replies": [] }, { "id": 741422, "author_name": "davidbnn92", "author_url": "", "post_date": "02/10/2020 16:02:54", "content": "I pictured you writing that with blood on the walls of a prison cell. I'll take the bait: is there a logic behind it? Like algebraic stacks or qabalah or something.
", "votes": null, "replies": [] }, { "id": 741430, "author_name": "scirpus", "author_url": "", "post_date": "02/10/2020 16:12:20", "content": "It is a tool I wrote that you give it a target and then it uses darwinian evolution to get to a solution - this took over 2000 generations. It generates latex, r, c++ and python.
", "votes": null, "replies": [] }, { "id": 743115, "author_name": "pfichou", "author_url": "", "post_date": "02/11/2020 19:53:09", "content": "\n\n« this took over 2000 generations »
\n\nMister Darwin I presume ...
\n\nI would like to understand your model(s), really !
", "votes": null, "replies": [] } ] }, { "id": 743111, "author_name": "pfichou", "author_url": "", "post_date": "02/11/2020 19:49:26", "content": "\n\nI really appreciate your contributions but, I still don’t understand Genetic.
\n\nI did ask you sometime ago some documentation, paper and I still would appreciate.
\n\nYour scripts are so obscure but efficient ;-)
\n\nRegards
", "votes": null, "replies": [ { "id": 744214, "author_name": "scirpus", "author_url": "", "post_date": "02/12/2020 16:57:50", "content": "Hey there - I did give you a pdf that expresses the code as pure math ;)
", "votes": null, "replies": [] } ] }, { "id": 744353, "author_name": "pfichou", "author_url": "", "post_date": "02/12/2020 19:01:08", "content": "Really ?
\n\nSomething a « standard »human could understand ?
\n\nIt is anyway fascinating :-)
", "votes": null, "replies": [] }, { "id": 746267, "author_name": "puremath86", "author_url": "", "post_date": "02/14/2020 19:51:35", "content": "I like the idea of embedding the \"missing\" values on the purely imaginary axis. Have you ever played around with sending them to different values
\n\n$$\n( 0+b_k \\cdot i ) \n$$
\n\nfor different features
\n\n$$\nX_k \\textrm{?}\n$$
", "votes": null, "replies": [ { "id": 746273, "author_name": "scirpus", "author_url": "", "post_date": "02/14/2020 19:54:40", "content": "I haven't yet - the code is pretty new - I'll keep you posted!
", "votes": null, "replies": [] }, { "id": 746290, "author_name": "puremath86", "author_url": "", "post_date": "02/14/2020 20:08:44", "content": "Awesome! Much appreciated.
\n\nOne further question: Without giving away your secret sauce --are you using a memetic algorithm per chance?
", "votes": null, "replies": [] }, { "id": 746314, "author_name": "scirpus", "author_url": "", "post_date": "02/14/2020 20:32:23", "content": "Straight Genetic Programming - Look at anything by John Koza - who also invented the lottery scratchcard!
", "votes": null, "replies": [] }, { "id": 746319, "author_name": "puremath86", "author_url": "", "post_date": "02/14/2020 20:38:38", "content": "Gotcha. Thanks for the reference!
", "votes": null, "replies": [] } ] }, { "id": 746278, "author_name": "pfichou", "author_url": "", "post_date": "02/14/2020 19:56:12", "content": "\n\nI wanna be among them too...
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "740634": "Here is a document detailing the complex math of one of the models given in https://www.kaggle.com/scirpus/handle-missing-values-not.\nHope you like it.\nn.b. The overlines are there to remind one that these are complex numbers!", "740753": "So edgy!", "740771": "More like so bored waiting for a non deep learning competition.\nI do Kaggle as a bit of fun rather than wanting to download an exabyte of image data ;)", "741422": "I pictured you writing that with blood on the walls of a prison cell. I'll take the bait: is there a logic behind it? Like algebraic stacks or qabalah or something.", "741430": "It is a tool I wrote that you give it a target and then it uses darwinian evolution to get to a solution - this took over 2000 generations. It generates latex, r, c++ and python.", "743111": "scirpus \n\nI really appreciate your contributions but, I still don’t understand Genetic.\n\nI did ask you sometime ago some documentation, paper and I still would appreciate.\n\nYour scripts are so obscure but efficient ;-)\n\nRegards", "743115": "scirpus \n\n« this took over 2000 generations » \n\nMister Darwin I presume ...\n\nI would like to understand your model(s), really !", "744214": "Hey there - I did give you a pdf that expresses the code as pure math ;)", "744353": "Really ?\n\nSomething a « standard »human could understand ?\n\nIt is anyway fascinating :-)", "746267": "I like the idea of embedding the \"missing\" values on the purely imaginary axis. Have you ever played around with sending them to different values \n\n$$\n( 0+b_k \\cdot i ) \n$$ \n\nfor different features \n\n$$\nX_k \\textrm{?}\n$$", "746273": "I haven't yet - the code is pretty new - I'll keep you posted!", "746278": "scirpus \n\nI wanna be among them too...", "746290": "Awesome! Much appreciated.\n\nOne further question: Without giving away your secret sauce --are you using a memetic algorithm per chance?", "746314": "Straight Genetic Programming - Look at anything by John Koza - who also invented the lottery scratchcard!", "746319": "Gotcha. Thanks for the reference!" }, "source": "meta" }