# Data science and prediction models with R

Data science and prediction models with R

## Overview

R started as a statistical programming tool, as the open source version of S. Unlike SPSS it requires some programming skills, since it is no drag and drop tool.

The CRAN repository holds numerous R packages where you can find several prepackaged functions which can turn out to be handy in many situations.

This course is the second part of our R track. It will handle the data science process and building prediction models and recommendation engines in R.

The course will give you hands on knowledge of these techniques in R, it will however not offer you a huge mathematical insight in the techniques used, therefor we refer to the course “The math behind data science” in our data science track.

You will very quickly learn the entire data science process, since it is the setting of prediction models and recommendation engines.

Furthermore, it is the second of three parts in our R track: in order to get started with R we refer to “Getting started programming with R”, for even more advanced stuff, we refer to “Building data products with R”.

Essentially, this course gives you a head start on data science with R

- The data science process: collect, describe, discover, predict, advise

- Recommendation engines

- Basic prediction models

- Enhanced/upgraded prediction models

- Deep learning techniques with R

Hands on exercises on all topics are offered

Learning objectives:

- Understanding the data science process and knowing where predictions fit in

- Make recommendations in R

- Discover the CRAN repository

- Learn how to use predefined machine learning techniques

- Implement your own machine learning technique

- Get an idea of how to improve your models, either by improving your data quality, your model parameters or your model itself

- Understand the basics of building a neural network with R

## Topics

CHAPTER 1: The data science process in R

- Collect

- Describe

- Discover

- Predict

- Advise

- Hands on example

CHAPTER 2: The CRAN repository

- Machine learning on CRAN

- Hands on with packages: the help function

CHAPTER 3: Recommendation engines

- Collaborative filtering techniques

--- Item based

--- User based

- Content-based filtering

--- Clustering

---Classifiers

CHAPTER 4: Machine learning techniques

- Class probability

- Classifiers

- Predefined machine learning techniques

- Create your own machine learning script

CHAPTER 5: Upgrading your model

- GAM

- Random Forest

- Support vector machine

CHAPTER 6: Deep learning with R

- Building a neural network

## Prerequisites

Unlike the drag and drop tool SPSS is R a programming tool. Basic R programming skills are required. If necessary, we refer to the course “Getting started with programming in R”, the predecessor of this course in the R track.

## Audience

This course is aimed towards management/BI personnel willing to build recommendation engines and prediction models in a user-friendly environment.

It is also aimed towards developers who are in need to employ these models for their management/BI environment.