Machine Learning Basics

Introduction

Machine learning in R uses various packages for predictive modeling. The tidymodels framework provides a consistent interface.

Basic Workflow

library(tidymodels)

# Split data
split <- initial_split(df, prop = 0.8)
train_data <- training(split)
test_data <- testing(split)

# Preprocess
recipe <- recipe(target ~ ., data = train_data) %>%
  step_normalize(all_numeric())

# Define model
model <- linear_reg() %>%
  set_engine("lm")

# Fit
workflow() %>%
  add_recipe(recipe) %>%
  add_model(model) %>%
  fit(train_data)

# Predict
predict(fit, test_data)

Model Types

# Linear regression
linear_reg()

# Logistic regression
logistic_reg()

# Decision tree
decision_tree()

# Random forest
rand_forest()

# K-nearest neighbors
nearest_neighbor()

Summary

tidymodels provides consistent ML workflow. Use appropriate model for your problem type.