Data Science with R
Artificial Intelligence (AI)
COURSE OVERVIEW
This comprehensive 6-day program is designed to transform participants into proficient data practitioners using the R programming language. Moving beyond legacy R practices, this course is built entirely on the Tidyverse and Tidymodels ecosystems—the modern industry standard for reproducible and scalable data science.
Participants will navigate the full data lifecycle: from complex data ingestion and "wrangling" to advanced statistical modeling and interactive communication. By Day 6, you will have moved from a beginner level to being able to architect a complete data product, including a predictive model and an interactive dashboard.
COURSE OBJECTIVES
By the end of this 6-day intensive, participants will be able to:
- Architect Tidy Workflows: Master the "Grammar of Data Manipulation" using dplyr and tidyr.
- Visualize Complex Insights: Build multi-layered, publication-quality graphics using ggplot2.
- Implement Modern Modeling: Use the tidymodels suite to build, cross-validate, and evaluate machine learning models.
- Automate Reporting: Create dynamic, reproducible documents using Quarto (the successor to R Markdown).
- Develop Interactive Apps: Build and deploy web-based data dashboards using Shiny.
- Handle Specialized Data: Manage dates/times with lubridate, strings with stringr, and functional programming with purrr.
Duration: 6 Days / 48 Hours
Delivery Method: Classroom-based, Virtual Instructor Led Training
COURSE OUTLINE
Day 1: Foundations & The Tidyverse Mindset
- R & RStudio for 2026: Setting up a modern development environment and Quarto projects.
- The Tidy Data Philosophy: Understanding tibbles and why "tidy" data is the key to efficiency.
- Data Ingestion: Loading data from CSV, Excel, SQL databases, and APIs (readr, haven, httr2).
- Hands-on: Importing a "messy" multi-source dataset and performing initial structural audits.
Day 2: Advanced Data Wrangling
- The Five Verbs of dplyr: filter, select, mutate, summarize, and arrange.
- Relational Data: Mastering joins (left_join, inner_join) to combine disparate data tables.
- Pivoting & Reshaping: Moving between "Wide" and "Long" formats with tidyr.
- Hands-on: Transforming raw transaction logs into a clean, customer-centric analytical base table.
Day 3: The Grammar of Graphics & EDA
- Visualizing with ggplot2: Mapping aesthetics, utilizing geoms, and mastering facets.
- Statistical EDA: Using visualization to spot outliers, skewness, and multi-collinearity.
- Publication-Ready Themes: Customizing scales, labels, and color palettes for executive presentations.
- Hands-on: Conducting a deep-dive Exploratory Data Analysis (EDA) on a global socioeconomic dataset.
Day 4: Statistical Inference & Linear Modeling
- Modern Inference: Using the infer package for hypothesis testing and confidence intervals.
- Regression Frameworks: Building linear and logistic models with the parsnip package.
- Feature Engineering: Creating "Recipes" for scaling, dummy encoding, and interaction terms.
- Hands-on: Building a predictive model to forecast real estate prices or customer churn.
Day 5: Machine Learning with Tidymodels
- The Tidymodels Ecosystem: Integrating rsample for cross-validation and tune for hyperparameter optimization.
- Advanced Algorithms: Deploying Random Forests and XGBoost via a unified interface.
- Model Evaluation: Using yardstick to calculate AUC, RMSE, and F1-scores across folds.
- Hands-on: Training and tuning a high-performance ensemble model to solve a classification challenge.
Day 6: Communication, Dashboards & Deployment
- Reproducible Reporting: Generating HTML and PDF reports using Quarto.
- Interactive Data Apps: Building a basic Shiny dashboard to allow users to filter and explore data.
- Model Interpretability: Using vip to explain feature importance and "Variable Importance."
- Capstone Presentation: Deploying an end-to-end data product: a reproducible report and an interactive predictor.
REGISTER NOW
