Machine Learning met R

De cursist leert de belangrijkste terminologie binnen de Machine Learning (ML) en leert de R libraries toe te passen bij machine learning.

Machine Learning is een onderdeel van artificial intelligence en gaat over

algoritmes die zich automatisch verbeteren op basis van ervaring. Deze Machine learning algoritmes maken een model gebaseerd op training data en kunnen

voorspellingen doen over nieuwe data.

Dag 1: Introductie dag. Korte herhaling R programmeren
De R Data Types en Data Frames
Statistische functies in R
R Data Files
R Packages
R Algoritmen

Daarna gaan we aan de slag met Machine Learning.

Wat is Machine Learning
Building Models of Data
Model Based Learning
Tunable Parameters
Supervised Learning
Discrete Labels
Continuous Labels
Classification and Regression
Unsupervised Learning
Clustering and Dimensionality Reduction Dag 2: Regression. Check Model
Using Summary
Using Coefficients
Correlation R
R Squared
F Test
Check Model Graphically
Check Residuals
Polynomial Regression
Gaussian Basis Functions
Overfitting
Compare with Linear Regression
Explore with Graphics
Logistic Function
Checking Model
Using Summary
Using Coefficients
Calculate Probabilities
Making Predictions
Confusion Matrix
Accuracy
Precision and Recall
ROC Curve
Functional R
Solving Iteration
purr package
library tidyverse
map Functions
Parameters of map
.x as placeholder
map_lgl Function
map_int and map_char
map2 Function
Other iteration functions
Combine purr with dyplr
walk Function Dag 3: Machine Learning. MSpark Session
Copy data into Spark
File Setup
Load data
Spark SQL
Store Data
Using dplyr
showquery()
Spark DataFrame Functions
sdf_pivot()
Feature Transformers
Distributed R
Web Applications
Shiny Architecture
Shiny Server
UI and Server
Input Object
Output Object
Reactivity
Render Options
Shiny Functions
Shiny Layout and Dashboard
Shiny Performance
Ensemble Learner
Creating Decision Trees
DecisionTreeClassifier
Overfitting Decision Trees
Ensembles of Estimator
Random Forests
Parallel Estimators
Bagging Classifier
Random Forest Regression
RandomForestRegressor
Non Parametric Model

- U bent in staat de zelflerende machine te trainen en toe te passen.
- U bent in staat om zelf een zelflerende machine te ontwikkelen.