tag:blogger.com,1999:blog-5952320191615496730.post1771004274957776888..comments2018-06-07T05:22:26.248+01:00Comments on The Beginner Programmer: Plain vanilla recurrent neural networks in R: waves predictionMichttp://www.blogger.com/profile/18151225177833588981noreply@blogger.comBlogger12125tag:blogger.com,1999:blog-5952320191615496730.post-3224609850761564272018-01-24T17:27:10.374+00:002018-01-24T17:27:10.374+00:00I have the same problem. Have you figured out how ...I have the same problem. Have you figured out how to solve it?Andrea Buccihttps://www.blogger.com/profile/01799678055535086261noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-36140693896375688252017-01-25T10:21:49.419+00:002017-01-25T10:21:49.419+00:00Hi, I'm afraid I can't really help you wit...Hi, I'm afraid I can't really help you with your specific problem, but you can ask on StackExchange's crossvalidated website where there are plenty of knowledgeable people ready to answer you questions.Michttps://www.blogger.com/profile/18151225177833588981noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-7184797454305094402017-01-16T14:02:25.136+00:002017-01-16T14:02:25.136+00:00Hello!
I'm trying to use this package, which s...Hello!<br />I'm trying to use this package, which seems to have anything I need for a time series prediction, still I'm stuck at the beginning - what's the correct way of presenting data to the rnn?<br />My dataset is made of:<br />Y (outcome variable)<br />Week (1-52, week of the year)<br />X1-X3 (indipendent variables)<br /><br />So I have Y = f(week, X1,X2,X3)<br /><br />According to the docs, <br /><br />"Y <br />array of output values, dim 1: samples (must be equal to dim 1 of X), dim 2: time (must be equal to dim 2 of X), dim 3: variables (could be 1 or more, if a matrix, will be coerce to array)<br />X <br />array of input values, dim 1: samples, dim 2: time, dim 3: variables (could be 1 or more, if a matrix, will be coerce to array)"<br /><br />But I can't quite figure it out!<br />Thanks in advance for your time!Zorba il Grecohttps://www.blogger.com/profile/17331326393251368604noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-715589892457244782016-12-02T15:36:00.622+00:002016-12-02T15:36:00.622+00:00Take a look at http://www.financial-hacker.com/bui...Take a look at http://www.financial-hacker.com/build-better-strategies-part-5-developing-a-machine-learning-system/Unknownhttps://www.blogger.com/profile/08280550194412808354noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-74076420864916002432016-12-02T15:34:17.913+00:002016-12-02T15:34:17.913+00:00This comment has been removed by the author.Unknownhttps://www.blogger.com/profile/08280550194412808354noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-58822417509450801012016-12-02T13:39:41.079+00:002016-12-02T13:39:41.079+00:00This comment has been removed by the author.Kanime H.https://www.blogger.com/profile/00755229091505122116noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-12674914308416760242016-12-02T12:36:40.391+00:002016-12-02T12:36:40.391+00:00Hi Kanime, take a look at https://stats.stackexcha...Hi Kanime, take a look at https://stats.stackexchange.com/ there's a community of people that may be able to help you with your questions. The people over there are very nice, helpful and skilled. If you look through the questions you might even find the very one you were about to ask (that happens to me very often).Michttps://www.blogger.com/profile/18151225177833588981noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-67145200608378540722016-12-01T23:42:39.376+00:002016-12-01T23:42:39.376+00:00This comment has been removed by the author.Kanime H.https://www.blogger.com/profile/00755229091505122116noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-45916620290753086882016-11-01T21:13:16.040+00:002016-11-01T21:13:16.040+00:00Maybe in the future I'll try using it and I...Maybe in the future I'll try using it and I'll make a post about it.<br /><br />To be honest though, in mosts of the tests I made with time series, RNN or a simple deep neural network seemed to work fine straight out of the box without that much fine tuning. If you do not have any specific reason to use LSTM you can try these first and see if you can get good results.Michttps://www.blogger.com/profile/18151225177833588981noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-60128152609653970622016-10-31T12:20:44.989+00:002016-10-31T12:20:44.989+00:00Thanks, I saw it already, but can’t figure out exa...Thanks, I saw it already, but can’t figure out exactly how to run time-series forecasting with MXNet and LTSM…<br />Unknownhttps://www.blogger.com/profile/08280550194412808354noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-41901601322379017012016-10-31T11:26:28.072+00:002016-10-31T11:26:28.072+00:00Hi! As far as I know, MXNet provides the LSTM mode...Hi! As far as I know, MXNet provides the LSTM model implementation. You can find some examples in the official documentation here: http://mxnet.io/tutorials/nlp/rnn.htmlMichttps://www.blogger.com/profile/18151225177833588981noreply@blogger.comtag:blogger.com,1999:blog-5952320191615496730.post-67011182280624608692016-10-30T15:58:12.036+00:002016-10-30T15:58:12.036+00:00I'm curious, how to implement the same code us...I'm curious, how to implement the same code using MXNET?Unknownhttps://www.blogger.com/profile/08280550194412808354noreply@blogger.com