Although very unusual, CNNs can also be used in time series classification. I made a very naive preprocessing and algorithm and there are lots of ways to improve it in case someone wants to give it a shot or at least bring some discussion about the use of this type of NN in this competition. Here is the kernel :)
", "messages": [ { "id": 422844, "postDate": "2018-11-16T21:37:18.813Z", "content": "Although very unusual, CNNs can also be used in time series classification. I made a very naive preprocessing and algorithm and there are lots of ways to improve it in case someone wants to give it a shot or at least bring some discussion about the use of this type of NN in this competition. Here is the kernel :)
", "rawMarkdown": "Although very unusual, CNNs can also be used in time series classification. I made a very naive preprocessing and algorithm and there are lots of ways to improve it in case someone wants to give it a shot or at least bring some discussion about the use of this type of NN in this competition. [Here](https://www.kaggle.com/hrmello/using-cnns-in-time-series-analysis/) is the kernel :) ", "votes": 5 }, { "id": 422930, "postDate": "2018-11-17T05:12:28.660Z", "content": "I am doing that, not as images but as 1D convolutions along each passband as channels. The results are similar to using GRU. (maybe I should tune it well)
", "rawMarkdown": "I am doing that, not as images but as 1D convolutions along each passband as channels. The results are similar to using GRU. (maybe I should tune it well)", "votes": 3, "replies": [ { "id": 423118, "postDate": "2018-11-17T14:28:48.883Z", "content": "Nice! Are you following some idea from a paper or blog post?
", "rawMarkdown": "Nice! Are you following some idea from a paper or blog post? " }, { "id": 423143, "postDate": "2018-11-17T15:55:06.593Z", "content": "No, I did not look for any papers. I just got the idea and implemented it in pytorch.
", "rawMarkdown": "No, I did not look for any papers. I just got the idea and implemented it in pytorch. " } ] }, { "id": 423125, "postDate": "2018-11-17T14:58:18.573Z", "content": "hello, I want to do classification to the lung cancer data by using convolutional Neural Network. can you help me. If you have a matlab code for this purpose, please send me
", "rawMarkdown": "hello, I want to do classification to the lung cancer data by using convolutional Neural Network. can you help me. If you have a matlab code for this purpose, please send me ", "votes": -3 }, { "id": 1990730, "postDate": "2022-10-16T18:00:09.910Z", "content": "hello, Can you please suggest me that how to apply CNN for the network traffic classification. Network traffic dataset is in the form of pcap file.
", "rawMarkdown": "hello, Can you please suggest me that how to apply CNN for the network traffic classification. Network traffic dataset is in the form of pcap file." }, { "id": 423507, "postDate": "2018-11-18T12:38:58.730Z", "content": "I am using 18 pixels wide band 10943292 pixels long, turned into a cylinder on which an ensemble of convolution matrixes slide around. I have shrunk it to 1024 pixels long, because computer said nooo...
", "rawMarkdown": "I am using 18 pixels wide band 10943292 pixels long, turned into a cylinder on which an ensemble of convolution matrixes slide around. I have shrunk it to 1024 pixels long, because computer said nooo..." }, { "id": 423177, "postDate": "2018-11-17T17:10:44.033Z", "rawMarkdown": "", "votes": 2, "isDeleted": true } ], "comments": [ { "id": 422930, "author_name": "Bharat", "author_url": "", "post_date": "2018-11-17T05:12:28.660000", "content": "I am doing that, not as images but as 1D convolutions along each passband as channels. The results are similar to using GRU. (maybe I should tune it well)
", "votes": 3, "replies": [ { "id": 423118, "author_name": "Henrique Mello", "author_url": "", "post_date": "2018-11-17T14:28:48.883000", "content": "Nice! Are you following some idea from a paper or blog post?
", "votes": 0, "replies": [] }, { "id": 423143, "author_name": "Bharat", "author_url": "", "post_date": "2018-11-17T15:55:06.593000", "content": "No, I did not look for any papers. I just got the idea and implemented it in pytorch.
", "votes": 0, "replies": [] } ] }, { "id": 423125, "author_name": "Mahmoud Abouagwa", "author_url": "", "post_date": "2018-11-17T14:58:18.573000", "content": "hello, I want to do classification to the lung cancer data by using convolutional Neural Network. can you help me. If you have a matlab code for this purpose, please send me
", "votes": -3, "replies": [] }, { "id": 1990730, "author_name": "Mirarani Choudhury", "author_url": "", "post_date": "2022-10-16T18:00:09.910000", "content": "hello, Can you please suggest me that how to apply CNN for the network traffic classification. Network traffic dataset is in the form of pcap file.
", "votes": 0, "replies": [] }, { "id": 423507, "author_name": "Tord Malmgren", "author_url": "", "post_date": "2018-11-18T12:38:58.730000", "content": "I am using 18 pixels wide band 10943292 pixels long, turned into a cylinder on which an ensemble of convolution matrixes slide around. I have shrunk it to 1024 pixels long, because computer said nooo...
", "votes": 0, "replies": [] }, { "id": 423177, "author_name": "", "author_url": "", "post_date": "2018-11-17T17:10:44.033000", "content": "", "votes": 2, "replies": [] } ], "raw_markdown_by_id": { "422844": "Although very unusual, CNNs can also be used in time series classification. I made a very naive preprocessing and algorithm and there are lots of ways to improve it in case someone wants to give it a shot or at least bring some discussion about the use of this type of NN in this competition. [Here](https://www.kaggle.com/hrmello/using-cnns-in-time-series-analysis/) is the kernel :) ", "422930": "I am doing that, not as images but as 1D convolutions along each passband as channels. The results are similar to using GRU. (maybe I should tune it well)", "423125": "hello, I want to do classification to the lung cancer data by using convolutional Neural Network. can you help me. If you have a matlab code for this purpose, please send me ", "1990730": "hello, Can you please suggest me that how to apply CNN for the network traffic classification. Network traffic dataset is in the form of pcap file.", "423507": "I am using 18 pixels wide band 10943292 pixels long, turned into a cylinder on which an ensemble of convolution matrixes slide around. I have shrunk it to 1024 pixels long, because computer said nooo...", "423177": "" } }