Machine Learning in Data Science

2019 CQS Summer Short Course: Machine Learning in Data Science

This CQS Summer Institute course provides an introduction and overview of the most powerful and common machine learning methods, includgin deep learning with neural networks, using R (The R Project for Statistical Computing: https://www.r-project.org). Upon completion of this course, participants will be able to select and apply off-the-shelf machine learning and statistical software tools to many different types of analytical tasks. This course requires some familiarity with R/RStudio and basic math.

Syllabus

Day 1: Data Management

• R and RStudio basics
  • RStudio IDE
• R data structures
  • Vectors
  • Lists
  • Data frames
• Importing and exporting data
  • Simulating data
  • Flat files (e.g., CSV)
  • RData files
  • External databases
• Data manipulation
  • reshape
  • melt and cast
  • subset

Day 2: Supervised Learning Part 1

• Supervised learning principles
  • Bias/variance trade-off
  • Objective (loss) functions
  • Optimism
  • Cross-validation
• Linear regression vs. K-nearest-neighbors
• Classification and regression trees
• Random forests (bagging)

Day 3: Supervised Learning Part 2

• Gradient boosted trees (boosting)
• Neural networks
  • Fully connected
  • Convolutional

Day 4: Supervised Learning Part 3

• Neural networks
  • Deep learning

Day 5: Unsupervised Learning

• Unsupervised learning principles
  • Model selection and parsimony
• Dimension reduction
  • Principal components analysis
• Clustering
  • K-means
  • Hierarchical

Background and Software Installation

This course requires the use of RStudio. This software is open source and freely available for download and installation on all operating systems. See https://www.rstudio.org/

We will additionally make use of several freely available add-on packages for R, which are listed below and can be installed using the following R commands:

install.packages("reshape2")

The neural network and deep learning topics will make use of the Keras library for R. Detailed information about getting started with this package is available at https://keras.rstudio.org/

This course is very ``hands on.'' I recommend that participants who are not familiar with R or RStudio work through some of the excellent and free online beginner tutorials for R/RStudio. The following link is a great place to start: https://support.rstudio.com/hc/en-us/articles/201141096-Getting-Started-with-R

Other Resources

• Introduction to Statistical Learning: http://www-bcf.usc.edu/~gareth/ISL/