LEARN COMPLETE PYTHON IN 24 HOURS

8. Machine Learning with R

R has excellent support for machine learning — especially for statistical modeling, interpretable models, and research-oriented workflows. In 2026, two main frameworks dominate: caret (classic, still widely used) and tidymodels (modern, tidyverse-integrated, recommended for new projects).

8.1 Supervised Learning – Regression & Classification

Supervised learning = predict an outcome variable (label) from input features.

• Regression → continuous target (price, temperature, sales)

• Classification → categorical target (yes/no, spam/not-spam, 0/1/2)

Common algorithms in R

• Linear / Logistic Regression

• Decision Trees & Random Forest

• Gradient Boosting (XGBoost, LightGBM, CatBoost)

• Support Vector Machines

• k-Nearest Neighbors

8.2 caret vs tidymodels – Two Main ML Frameworks

Featurecaret (older, still very popular)tidymodels (modern, tidyverse-style)Winner in 2026SyntaxFunctional, base-R styleConsistent, pipe-friendly, tidyverse ecosystemtidymodelsPreprocessingBuilt-in preProcess()recipes package (very powerful)tidymodelsModel tuningtrain() with gridtune_grid(), tune_bayes()tidymodelsWorkflowManual stepsworkflow() – combines recipe + modeltidymodelsCommunity momentumLarge legacy user baseRapidly growing, Posit-supportedtidymodelsLearning curveModerateSlightly steeper at first, then easier—

Recommendation (2026): Use tidymodels for all new work — it’s more readable, reproducible, and integrates perfectly with tidyverse. Learn caret only if maintaining legacy code.

Quick tidymodels example

R

library(tidymodels) # Split data set.seed(42) split <- initial_split(iris, prop = 0.8, strata = Species) train_data <- training(split) test_data <- testing(split) # Recipe (preprocessing) rec <- recipe(Species ~ ., data = train_data) %>% step_normalize(all_numeric_predictors()) # Model rf_model <- rand_forest(trees = 500) %>% set_mode("classification") %>% set_engine("ranger") # Workflow wf <- workflow() %>% add_recipe(rec) %>% add_model(rf_model) # Fit fit <- wf %>% fit(data = train_data) # Predict & evaluate predictions <- predict(fit, test_data) accuracy <- accuracy_vec(test_data$Species, predictions$.pred_class) print(accuracy)

8.3 Random Forest, XGBoost & Gradient Boosting in R

Random Forest (bagging ensemble – very robust)

R

# tidymodels way rf_spec <- rand_forest(trees = tune(), min_n = tune()) %>% set_mode("regression") %>% set_engine("ranger") # Tune tune_res <- tune_grid( rf_spec, mpg ~ ., resamples = vfold_cv(mtcars, v = 5), grid = 10 ) best_params <- select_best(tune_res, "rmse") final_model <- finalize_model(rf_spec, best_params)

XGBoost (gradient boosting – often top performer)

R

# Install: install.packages("xgboost") library(xgboost) # Prepare data (matrix format required) X <- as.matrix(mtcars[, -1]) y <- mtcars$mpg xgb_model <- xgboost( data = X, label = y, nrounds = 100, objective = "reg:squarederror", eta = 0.1, max_depth = 6 ) # Prediction pred <- predict(xgb_model, X) rmse <- sqrt(mean((y - pred)^2)) print(rmse)

Gradient Boosting comparison (2026)

• XGBoost → fastest, most accurate, GPU support

• LightGBM → even faster on large data

• CatBoost → best for categorical features out-of-the-box

8.4 Model Evaluation – Cross-validation, ROC-AUC, Confusion Matrix

Cross-validation in tidymodels

R

folds <- vfold_cv(train_data, v = 10, strata = Species) metrics <- metric_set(accuracy, roc_auc) res <- fit_resamples( rf_model, resamples = folds, metrics = metrics ) collect_metrics(res)

Confusion Matrix & ROC-AUC

R

# Classification example confusion <- conf_mat(predictions, truth = test_data$Species) confusion %>% autoplot(type = "heatmap") # ROC-AUC (binary classification) roc_auc_vec(truth = y_test, estimate = prob_positive_class)

Regression metrics

• RMSE, MAE, R² (rsq_trad)

8.5 Unsupervised Learning – Clustering (k-means, hierarchical)

K-Means Clustering

R

# Scale data first! data_scaled <- scale(mtcars[, c("mpg", "hp", "wt")]) # K-means km <- kmeans(data_scaled, centers = 3, nstart = 25) # Visualize mtcars$cluster <- factor(km$cluster) ggplot(mtcars, aes(x = mpg, y = hp, color = cluster)) + geom_point(size = 4) + labs(title = "K-Means Clustering – mtcars")

Hierarchical Clustering

R

dist_matrix <- dist(data_scaled, method = "euclidean") hc <- hclust(dist_matrix, method = "complete") plot(hc, main = "Hierarchical Clustering Dendrogram") rect.hclust(hc, k = 3, border = "red")

Choosing k (number of clusters)

R

library(factoextra) fviz_nbclust(data_scaled, kmeans, method = "wss") + labs(title = "Elbow Method") fviz_nbclust(data_scaled, kmeans, method = "silhouette") # Silhouette score

Mini Summary Project – Customer Segmentation

R

library(tidyverse) # Load sample customer data (or your own) df <- read_csv("customer_data.csv") # Preprocess df_scaled <- df %>% select(age, annual_income, spending_score) %>% scale() # K-means k <- 5 km <- kmeans(df_scaled, centers = k, nstart = 25) df$segment <- factor(km$cluster) # Visualize ggplot(df, aes(x = annual_income, y = spending_score, color = segment)) + geom_point(size = 4) + labs(title = "Customer Segments", subtitle = "Based on Income & Spending Score")

This completes the full Machine Learning with R section — now you can build, evaluate, and deploy real ML models in R!

📚 Amazon Book Library

All my books are FREE on Amazon Kindle Unlimited🌍 Exclusive Country-Wise Amazon Book Library – Only Here!

On GlobalCodeMaster.com you’ll find complete, ready-to-use lists of my books with direct Amazon links for every country.

Belong to India, Australia, USA, UK, Canada or any other country? Just click your country’s link and enjoy:

✅ Any eBook FREE on Kindle Unlimited ✅ Or buy at incredibly low prices

400+ fresh books written in 2025-2026 with today’s latest AI, Python, Machine Learning & tech trends – nowhere else will you find this complete country-wise collection on one platform!

Choose your country below and start reading instantly 🚀

BOOK LIBRARY USA 2026 LINK

BOOK LIBRARY INDIA 2026 LINK

BOOK LIBRARY AUSTRALIA 2026 LINK

BOOK LIBRARY CANADA 2026 LINK

BOOK LIBRARY UNITED KINGDOM 2026 LINK

BOOK LIBRARY GERMANY 2026 LINK

BOOK LIBRARY FRANCE 2026 LINK

BOOK LIBRARY ITALY 2026 LINK

BOOK LIBRARY SPAIN 2026 LINK

BOOK LIBRARY NETHERLANDS 2026 LINK

BOOK LIBRARY BRAZIL 2026 LINK

BOOK LIBRARY MEXICO 2026 LINK

BOOK LIBRARY JAPAN 2026 LINK

BOOK LIBRARY POLAND 2026 LINK

BOOK LIBRARY IRELAND 2026 LINK

BOOK LIBRARY SWEDEN 2026 LINK

BOOK LIBRARY BELGIUM 2026 LINK