Home

login

linkedin

Help

About me

Github

Courses

Medium

First%20time%20writing%20on%20Machine%20learning.%20In%20this%20post%20will%20discuss%20about%20text%20classification.%20We%20will%20use%20Naive%20Bayes%20Classifier%20for%20classification.%0A%0ALets%20start%2C%0A%0AOk%2C%20You%20may%20want%20to%20know%20what%20is%20classifier%20%3F%0A%0AWell%2C%20classifier%20classifies%20training%20data%20based%20on%20implemented%20algorithm%20as%20we%20are%20using%20Naive%20Bayes%20Classifier%20algorithm%20which%20is%20based%20on%20Bayes%20theorem.%20Naive%20Bayes%20Classifier%20works%20based%20on%20phrase%20frequency%20and%20its%20Supervised%20learning.%0A%0ALets%20assume%20Nur%20started%20using%20email%20and%20his%20relatives%2C%20teachers%2C%20friends%20are%20contacting%20via%20it.%20As%20well%20as%20he%20is%20using%20this%20email%20id%20in%20different%20social%20networks.%20As%20he%20is%20getting%20lots%20of%20email%20now%2C%20he%20want%20to%20categorize%20it%20so%20he%20can%20easily%20check%20emails%20from%20relatives%2C%20teachers%2C%20friends%20and%20others.%0A%0A!%5B%5D(http%3A%2F%2Fblog.sakib.ninja%2Fcontent%2Fimages%2F2017%2F02%2FScreen-Shot-2017-02-14-at-7.42.53-PM.png)%0A%0ANur%20has%20a%20MagicBox%20which%20is%20able%20to%20identify%20certain%20things%20based%20on%20preset%20data.%20So%20nur%20decides%20to%20take%20the%20help%20of%20magic%20box.%20So%20he%20started%20setting%20few%20key%20sentence%20and%20word%20based%20on%20which%20magic%20box%20will%20categorize%20emails.%0A%0AKey%20sentence%20and%20words%20are%20like%2C%0A%0A%5C%5BText%5C%5D%20%3D%3E%20%5C%5BLabel%5C%5D%0A%0Ayour%20aunty%20%3D%3E%20%5C%5Brelative%5C%5D%0A%0Adear%20friend%20%3D%3E%20%5C%5Bfriend%5C%5D%0A%0Afrom%20Jhalokathi%20%3D%20%5C%5Brelative%5C%5D%0A%0Ayour%20cousin%20%3D%3E%20%5C%5Brelative%5C%5D%0A%0ADear%20subscriber%20%3D%3E%20%5C%5Bsocial%5C%5D%0A%0Anew%20post%20%3D%3E%20%5C%5Bsocial%5C%5D%0A%0Anew%20homework%20%3D%3E%20%5C%5Bteacher%5C%5D%0A%0Ais%20weekend%20%3D%3E%20%5C%5Bteacher%5C%5D%0A%0Aclass%20teacher%20%5C%5Bteacher%5C%5D%0A%0AIn%20machine%20learning%20this%20are%20called%20training%20data.%0A%0ANow%20assume%20Nur%20got%20two%20email%2C%20Those%20are%2C%0A%0A**Hello%20nur%2C%20I%20am%20your%20class%20teacher%20Jessica.%20Tomorrow%20is%20weekend.%20So%20there%20is%20no%20homework.%20Enjoy%20the%20weekend.**%0A%0AAnd%0A%0A**Hello%20nur%2C%20I%20am%20Rupa%20your%20cousin.**%0A%0ANow%20lets%20check%20email%20category%20using%20Magic%20Box.%0A%0A!%5B%5D(http%3A%2F%2Fblog.sakib.ninja%2Fcontent%2Fimages%2F2017%2F02%2FScreen-Shot-2017-02-14-at-7.59.23-PM.png)%0A%0A!%5B%5D(http%3A%2F%2Fblog.sakib.ninja%2Fcontent%2Fimages%2F2017%2F02%2FScreen-Shot-2017-02-14-at-7.58.50-PM.png)%0A%0AOk%2C%20Nur%20got%20results%20from%20Magic%20Box%20and%20its%20correct.%0A%0ANurs%20Magic%20Box%20uses%20Naive%20Bayes%20Classifier%20algorithm%20to%20categorize%20data.%0A%0AIn%20supervised%20machine%20learning%20result%20accuracy%20depends%20on%20training%20data.%0A%0AUsed%20implementation%20of%20Naive%20Bayes%20Classifier%20%3A%20%5Bhttps%3A%2F%2Fgithub.com%2Fptnplanet%2FJava-Naive-Bayes-Classifier%5D(https%3A%2F%2Fgithub.com%2Fptnplanet%2FJava-Naive-Bayes-Classifier)%0A%0AExample%20%3A%0A%0A%60%60%60%0Aval%20bayesClassifier%20%3D%20BayesClassifier%3CString%2C%20String%3E()%0A%0A%2F%2F%20Learning%0Aval%20tokenizeSampleText%20%3D%20%22Sample%20Sentence%22.split(%22%20%22)%0Aval%20sampleLabel%20%3D%20%22SampleLabel%22%0AbayesClassifier.learn(sampleLabel%2C%20tokenizeSampleText)%0A%0A%2F%2F%20Classify%0Aval%20result%20%3D%20bayesClassifier.classify(%22QueryText%22.split(%22%20%22))%0Aprintln(%22Category%20%3A%20%24%7Bresult.category%7D%2C%20Probability%20%3A%20%24%7Bresult.probability%7D%22)%0A%0A%60%60%60%0A%0AMagic%20Box%20source%20code%20%3A%0A%0A%5Bhttps%3A%2F%2Fgithub.com%2Fs4kibs4mi%2Fmachine-learning-garage%5D(https%3A%2F%2Fgithub.com%2Fs4kibs4mi%2Fmachine-learning-garage)

ML : Nur & His Magic Box

Bangladesh

Learn with Sakib