--- title: How to perform stepwise variable selection in logistic regression model? description: How to perform stepwise variable selection in logistic regression model? --- [Skip to content](https://help.datasciencedojo.com/migration/how-to-perform-stepwise-variable-selection-in-logistic-regression-model#main-content) English Show submenu for translations [Customer portal](https://help.datasciencedojo.com/tickets-view?hsLang=en) ![DSD main logo.png\]](https://help.datasciencedojo.com/hs-fs/hubfs/DSD%20main%20logo.png?height=40&name=DSD%20main%20logo.png) - [Default HubSpot Blog](https://help.datasciencedojo.com/blog) Open main navigation Close main navigation - [Default HubSpot Blog](https://help.datasciencedojo.com/blog) - English Show submenu for translations - [Customer portal](https://help.datasciencedojo.com/tickets-view) - Go to datasciencedojo.com Go to datasciencedojo.com How can we help? - There are no suggestions because the search field is empty. 1. [Help Center](https://help.datasciencedojo.com/migration?hsLang=en) 2. [R Language](https://help.datasciencedojo.com/migration/r-language?hsLang=en) # How to perform stepwise variable selection in logistic regression model? Preparing the data Data set: PimaIndiansDiabetes2 \[in mlbench package\], introduced in Chapter @ref(classification-in-r), for predicting the probability of being diabetes positive based on multiple clinical variables.   We’ll randomly split the data into training set (80% for building a predictive model) and test set (20% for evaluating the model). Make sure to set seed for reproductibility.   \# Load the data and remove NAs data("PimaIndiansDiabetes2", package = "mlbench") PimaIndiansDiabetes2 \<- na.omit(PimaIndiansDiabetes2) \# Inspect the data sample\_n(PimaIndiansDiabetes2, 3) \# Split the data into training and test set set.seed(123) training.samples \<- PimaIndiansDiabetes2$diabetes %\>% createDataPartition(p = 0.8, list = FALSE) train.data \<- PimaIndiansDiabetes2\[training.samples, \] test.data \<- PimaIndiansDiabetes2\[-training.samples, \] Computing stepwise logistique regression The stepwise logistic regression can be easily computed using the R function stepAIC() available in the MASS package. It performs model selection by AIC. It has an option called direction, which can have the following values: “both”, “forward”, “backward” (see Chapter @ref(stepwise-regression)).   Quick start R code library(MASS) \# Fit the model model \<- glm(diabetes ~., data = train.data, family = binomial) %\>% stepAIC(trace = FALSE) \# Summarize the final selected model summary(model) \# Make predictions probabilities \<- model %\>% predict(test.data, type = "response") predicted.classes \<- ifelse(probabilities \> 0.5, "pos", "neg") \# Model accuracy mean(predicted.classes==test.data$diabetes) Full logistic regression model Full model incorporating all predictors:   full.model \<- glm(diabetes ~., data = train.data, family = binomial) coef(full.model) \## (Intercept) pregnant glucose pressure triceps insulin \## -9.50372 0.04571 0.04230 -0.00700 0.01858 -0.00159 \## mass pedigree age \## 0.04502 0.96845 0.04256 Perform stepwise variable selection Select the most contributive variables:   library(MASS) step.model \<- full.model %\>% stepAIC(trace = FALSE) coef(step.model) \## (Intercept) glucose mass pedigree age \## -9.5612 0.0379 0.0523 0.9697 0.0529 The function chose a final model in which one variable has been removed from the original full model. Dropped predictor is: triceps.   Compare the full and the stepwise models Here, we’ll compare the performance of the full and the stepwise logistic models. The best model is defined as the model that has the lowest classification error rate in predicting the class of new test data:   Prediction accuracy of the full logistic regression model:   \# Make predictions probabilities \<- full.model %\>% predict(test.data, type = "response") predicted.classes \<- ifelse(probabilities \> 0.5, "pos", "neg") \# Prediction accuracy observed.classes \<- test.data$diabetes mean(predicted.classes == observed.classes) \## \[1\] 0.808 Prediction accuracy of the stepwise logistic regression model:   \# Make predictions probabilities \<- predict(step.model, test.data, type = "response") predicted.classes \<- ifelse(probabilities \> 0.5, "pos", "neg") \# Prediction accuracy observed.classes \<- test.data$diabetes mean(predicted.classes == observed.classes) \## \[1\] 0.795 - [Data Science Bootcamp](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#main-content) - [Certificate](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#certificate) - [Logistics](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#logistics) - [Curriculum](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#curriculum) - [Target Audience](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#target-audience) - [Talk to an Advisor](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#talk-to-an-advisor) - [Prerequisites](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#prerequisites) - [Pricing](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#pricing) - [Policies](https://help.datasciencedojo.com/migration/data-science-bootcamp?hsLang=en#policies) - [Large Language Models Bootcamp](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#main-content) - [Logistics](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#logistics) - [Curriculum](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#curriculum) - [Certificate](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#certificate) - [Prerequisites](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#prerequisites) - [Eligibility](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#eligibility) - [Talk to an advisor](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#talk-to-an-advisor) - [Pricing](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#pricing) - [Delivery](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#delivery) - [Program Overview](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#program-overview) - [Policies](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#policies) - [Registration](https://help.datasciencedojo.com/migration/large-language-models-bootcamp?hsLang=en#registration) - [Agentic AI Bootcamp](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#main-content) - [Prerequisites](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#prerequisites) - [Logistics](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#logistics) - [Curriculum](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#curriculum) - [Program Overview](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#program-overview) - [Pricing](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#pricing) - [Preparatory Material](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#preparatory-material) - [Policies](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#policies) - [Talk to an Advisor](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#talk-to-an-advisor) - [Registration](https://help.datasciencedojo.com/migration/agentic-ai-bootcamp?hsLang=en#registration) - [Registration](https://help.datasciencedojo.com/migration/registration?hsLang=en#main-content) - [Payment Methods](https://help.datasciencedojo.com/migration/registration?hsLang=en#payment-methods) - [Free Certificate](https://help.datasciencedojo.com/migration/registration?hsLang=en#free-certificate) - [Tuition Benefits](https://help.datasciencedojo.com/migration/registration?hsLang=en#tuition-benefits) - [Pricing](https://help.datasciencedojo.com/migration/pricing?hsLang=en) - [Community](https://help.datasciencedojo.com/migration/community?hsLang=en#main-content) - [FODAI 2025 - Sept Edition](https://help.datasciencedojo.com/migration/community?hsLang=en#fodai-2025-sept-edition) - [FODAI 2026](https://help.datasciencedojo.com/migration/community?hsLang=en#fodai-2026) - [Python Programming](https://help.datasciencedojo.com/migration/python-programming?hsLang=en) - [Platform Related Issues](https://help.datasciencedojo.com/migration/platform-related-issues?hsLang=en#main-content) - [Community](https://help.datasciencedojo.com/migration/platform-related-issues?hsLang=en#community) - [Bootcamps](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#main-content) - [Formats and Delivery](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#formats-and-delivery) - [Certificates](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#certificates) - [Policies](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#policies) - [Session Access](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#session-access) - [Benefits](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#benefits) - [Exercises](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#exercises) - [Tuition Benefits](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#tuition-benefits) - [Logistics](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#logistics) - [Testimonials](https://help.datasciencedojo.com/migration/bootcamps?hsLang=en#testimonials) - [Homework and Notebooks](https://help.datasciencedojo.com/migration/homework-and-notebooks?hsLang=en#main-content) - [Personal Assigments](https://help.datasciencedojo.com/migration/homework-and-notebooks?hsLang=en#personal-assigments) - [Free Courses](https://help.datasciencedojo.com/migration/free-courses?hsLang=en) - [Data Science for Business](https://help.datasciencedojo.com/migration/data-science-for-business?hsLang=en#main-content) - [Program Overview](https://help.datasciencedojo.com/migration/data-science-for-business?hsLang=en#program-overview) - [Practicum](https://help.datasciencedojo.com/migration/practicum?hsLang=en#main-content) - [Logistics](https://help.datasciencedojo.com/migration/practicum?hsLang=en#logistics) - [Program Overview](https://help.datasciencedojo.com/migration/practicum?hsLang=en#program-overview) - [Blog](https://help.datasciencedojo.com/migration/blog?hsLang=en#main-content) - [Content Collaboration](https://help.datasciencedojo.com/migration/blog?hsLang=en#content-collaboration) - [Guest Posts](https://help.datasciencedojo.com/migration/blog?hsLang=en#guest-posts) - [Link Exchanges/Insertions](https://help.datasciencedojo.com/migration/blog?hsLang=en#link-exchanges-insertions) - [Employment Assistance](https://help.datasciencedojo.com/migration/employment-assistance?hsLang=en#main-content) - [Resumes](https://help.datasciencedojo.com/migration/employment-assistance?hsLang=en#resumes) - [Internships](https://help.datasciencedojo.com/migration/employment-assistance?hsLang=en#internships) - [Profile Shortlisting](https://help.datasciencedojo.com/migration/employment-assistance?hsLang=en#profile-shortlisting) - [Career Paths](https://help.datasciencedojo.com/migration/employment-assistance?hsLang=en#career-paths) - [Applying for Mentor](https://help.datasciencedojo.com/migration/employment-assistance?hsLang=en#applying-for-mentor) - [Machine Learning](https://help.datasciencedojo.com/migration/machine-learning?hsLang=en#main-content) - [Unsupervised](https://help.datasciencedojo.com/migration/machine-learning?hsLang=en#unsupervised) - [Supervised](https://help.datasciencedojo.com/migration/machine-learning?hsLang=en#supervised) - [Statistics](https://help.datasciencedojo.com/migration/machine-learning?hsLang=en#statistics) - [Deep Learning](https://help.datasciencedojo.com/migration/machine-learning?hsLang=en#deep-learning) - [Timeseries](https://help.datasciencedojo.com/migration/machine-learning?hsLang=en#timeseries) - [Data Analysis](https://help.datasciencedojo.com/migration/data-analysis?hsLang=en) - [R Language](https://help.datasciencedojo.com/migration/r-language?hsLang=en) - [Python for Data Science](https://help.datasciencedojo.com/migration/python-for-data-science?hsLang=en#main-content) - [Prerequisites](https://help.datasciencedojo.com/migration/python-for-data-science?hsLang=en#prerequisites) - [Pricing](https://help.datasciencedojo.com/migration/python-for-data-science?hsLang=en#pricing) - [Curriculum](https://help.datasciencedojo.com/migration/python-for-data-science?hsLang=en#curriculum) - [Program Overview](https://help.datasciencedojo.com/migration/python-for-data-science?hsLang=en#program-overview) - [SQL](https://help.datasciencedojo.com/migration/sql?hsLang=en) - [Introduction to Power BI](https://help.datasciencedojo.com/migration/introduction-to-power-bi?hsLang=en) - [Power BI](https://help.datasciencedojo.com/migration/power-bi?hsLang=en) - [Programming and Tools](https://help.datasciencedojo.com/migration/programming-and-tools?hsLang=en#main-content) - [SQL](https://help.datasciencedojo.com/migration/programming-and-tools?hsLang=en#sql) - [Partnerships](https://help.datasciencedojo.com/migration/partnerships?hsLang=en) - [Default HubSpot Blog](https://help.datasciencedojo.com/blog) [![Chill listening crop-3](https://help.datasciencedojo.com/hs-fs/hubfs/DSD%20main%20logo.png?width=95&height=24&name=DSD%20main%20logo.png "Chill listening crop-3")](https://datasciencedojo.com) Copyright © 2026, Data Science Dojo