tag:blogger.com,1999:blog-7958828565254404797.post5702128783397440486..comments2024-03-18T23:27:20.269-07:00Comments on ListenData: Feature Selection : Select Important Variables with Boruta PackageDeepanshu Bhallahttp://www.blogger.com/profile/09802839558125192674noreply@blogger.comBlogger11125tag:blogger.com,1999:blog-7958828565254404797.post-41668942353698490252021-08-16T06:30:18.386-07:002021-08-16T06:30:18.386-07:00Nics post Thankyou for sharing...very helpful for ...Nics post Thankyou for sharing...very helpful for me..<br />Ishuhttps://www.bestbus.in/noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-74150927255126333212021-07-22T05:52:50.946-07:002021-07-22T05:52:50.946-07:00This comment has been removed by the author.Educational Serviceshttps://www.blogger.com/profile/12343449288195106852noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-36636060579647140722020-11-15T23:16:14.849-08:002020-11-15T23:16:14.849-08:00SIR CAN YOU HELP US WITH EDA PROCESS IN PHYTHONSIR CAN YOU HELP US WITH EDA PROCESS IN PHYTHONAnonymoushttps://www.blogger.com/profile/08305164239010450168noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-20350636650089098452020-06-16T02:56:58.908-07:002020-06-16T02:56:58.908-07:00This comment has been removed by a blog administrator.Create N Gifthttps://www.blogger.com/profile/11994049963243187894noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-6839010697170157232020-02-06T02:28:24.579-08:002020-02-06T02:28:24.579-08:00This comment has been removed by a blog administrator.Learn Bench Indiahttps://www.blogger.com/profile/13026128560145611224noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-6301517192675578092019-05-26T22:20:38.060-07:002019-05-26T22:20:38.060-07:00thank you sharing with us...
https://www.senseinte...thank you sharing with us...<br />https://www.senseinteriors.in/sensedhttps://www.senseinteriors.in/noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-76467187628789575662017-12-06T01:13:45.228-08:002017-12-06T01:13:45.228-08:00Hi Deepanshu,
How to implement this in SAS?
Rega...Hi Deepanshu,<br /><br />How to implement this in SAS?<br /><br />Regards,<br />Harneet.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-92135470484058835162017-11-15T06:10:32.408-08:002017-11-15T06:10:32.408-08:00Missing Value Imputation
Categorical logreg fo...Missing Value Imputation <br />Categorical logreg for Binary Variables , polyreg for more than 2 levels (MICE) <br />Numerical KNN Imputation , Multivariate Imputation by Chained Equations, rpart http://www.stefvanbuuren.nl/publications/MICE%20V1.0%20Manual%20TNO00038%202000.pdf<br /> <br />Levels Reduction WOE and Informaion and Business Logic <br />Dimensionality Reduction PCA <br />Data Synthesis SMOTE, ROSE . Synthetic Minority Oversampling Technique <br /> <br />Examining the Each variable outlier Detection, Outlier Capping for the numerical variables, bi-variate analysis wrt dependent variable <br />Test & Train Data preparation Caret package- createDataPartition <br />Model Building <br /> Logistic Regression <br /> stepAic and decision tree for the variable selection <br /> building the Logistic Regression model <br /> Evaluation of Logistic Regression <br /> Goodness of Fit -hoslem, loglikelihood, pR2,waldtest , variable Importance, Classification rate, AUC ROC curve, K-fold cross validation ,Concordance -Discordance pairs <br /> Evaluating the accuracy by random forest and CHAID trees <br /> Visulisation of the result via ggplots, <br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-1728835241255539822017-11-08T04:44:28.239-08:002017-11-08T04:44:28.239-08:00Awesome Post! Boruta is an easy to use package as ...Awesome Post! Boruta is an easy to use package as there aren’t many parameters to tune / remember. You shouldn’t use a data set with missing values to check important variables using Boruta.Naina@Digital Marketing Coursehttps://www.hiaim.conoreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-71668202519063169402017-11-06T22:20:35.076-08:002017-11-06T22:20:35.076-08:00TestIndex <- createDataPartition(Credit$Credit....TestIndex <- createDataPartition(Credit$Credit.Amount,times = 1,p=0.5,list = FALSE)<br />Train <- Credit[TestIndex,]<br />Test <- Credit[-TestIndex,]<br /><br />names(Credit)<br /># Building the model<br /><br />Fit <- glm(Creditability~Account.Balance+Payment.Status.of.Previous.Credit+Purpose+Value.Savings.Stocks+Length.of.current.employment+Type.of.apartment+Most.valuable.available.asset+Concurrent.Credits+Duration.of.Credit..month.+Credit.Amount+Age..years.,family = binomial,data = Train)<br /><br /><br />summary(Fit)<br /><br /># Removing the non significant variable<br /><br />Fit1 <- glm(Creditability~Account.Balance+Payment.Status.of.Previous.Credit+Value.Savings.Stocks+Length.of.current.employment+Most.valuable.available.asset+Duration.of.Credit..month.,family = binomial,data=Train )<br /><br /><br />summary(Fit1)<br /><br /><br />fitlog <- predict(Fit1,type="response",newdata=Test)<br />fitlog<br />fitlog1 <- predict(Fit1,type="response",newdata=Test)<br />fitlog1<-ifelse(fitlog1>0.5,1,0)<br /><br />tab <- table(fitlog1,Train$Creditability)<br /><br />1-sum(diag(tab))/sum(tab)<br /><br /># Model performance Evaluation<br />library(ROCR)<br />pred <- predict(Fit1,Test,type="response")<br />hist(pred)<br /><br /># Cut offvalue on eye estimate<br />pred <- prediction(pred,Test$Creditability)<br />eval <- performance(pred,"acc")<br />plot(eval)<br />abline(h=0.775,v=.575)<br /><br /><br /># Identifying the best cutoff and accuracy <br />eval<br />max <- which.max(slot(eval,"y.values")[[1]])<br />acc <- slot(eval,"y.values")[[1]][max]<br />cut <- slot(eval,"x.values")[[1]][max]<br />cut <br /><br /># Optimalcutoffvalueis .572<br />fitlog1 <- predict(Fit1,Test,type = "response")<br />fitlog1 <- ifelse(fitlog1>0.57,1,0)<br /><br /># Misclassification error <br />tab1 <- table(fitlog1,Test$Creditability)<br />tab1<br /><br />accuracy <- (sum(diag(tab1))/sum(tab1))<br />accuracy<br /># accuracy = 77.4%<br /><br /># ROC<br /># we are intrested in finding the number of 1 rather than 0's.'<br /><br />roc <- performance(pred,"tpr","fpr")<br />plot(roc)<br />abline(0,1)<br /><br />require(Deducer)<br />rocplot(Fit1)HARSHAhttps://www.blogger.com/profile/07186667650493020625noreply@blogger.comtag:blogger.com,1999:blog-7958828565254404797.post-62527557616782083612017-07-31T22:08:48.133-07:002017-07-31T22:08:48.133-07:00Very Useful data about this new trend topic, this ...Very Useful data about this new trend topic, this is good for each studying and expertise. And additionally these days I was attempting to find <a href="https://essayreviewratings.com/essay_review_reading_make_to_write_academic_essays_smoothly_and_cohesively.html" rel="nofollow">best essay writing service reviews</a> have quite a few right key points and I discover ways to write essays and get right of entry to it. I think which covers each and every points concerning with essay writing and which makes it greater powerful and beneficial reference for the me.Anonymoushttps://www.blogger.com/profile/00040149432315527017noreply@blogger.com