KEUZE SEMESTER BIG DATA. WHO AM I? Peter Odenhoven [email protected] Amsterdam University of Applied Sciences Background: Mathematics / Statistics.

KEUZE SEMESTER BIG DATA

WHO AM I?

• Peter Odenhoven• [email protected]

• Amsterdam University of Applied Sciences• Background: Mathematics / Statistics• Teaching at this moment:

• Programming: Java, C# , Python, R• Databases: SQL and NOSQL• Data warehousing / Business Intelligence • Big Data

IT’S ALL ABOUT FINDING STUFF …

PROJECT ASSIGNMENTS

1. HvA CHIEF: charging electrical cars

2. AFC AJAX : monotoring physics of player

3. CURVE FEVER: toxic behaviour in a multiplayer game

4. Digital Life Centre: sensor data

5. NIKHEF:CERN information use

10 REASONS WHY DATA SCIENTIST IS THE SEXIEST JOB OF THE 21ST CENTURY (OR NOT)

WHAT YOU SHOULD BRING,WHAT YOU SHOULD GET

NOT ALL PROBLEMS ARE BIG DATA PROBLEMS

OLD AND NEW FRIENDS: SQL AND NOSQL

module description ecProject Big data applications 61 Business Studio 42 Data analysis/mining 4 Business skills 1

total 15Project Big data applications 53 Data processing and storage 44 Information visualization 4 Professional skills 2

total 15

“… DO I LOOK LIKE A MAN WITH A PLAN…”

BUSINESS STUDIO

• Business opportunities• Data warehousing• Reporting• Ethics• Legal issues

DATA ANALYSIS AND DATA MINING

• Some mathematics, MAPLE• More theory on algoritms• A lot of practice:

• Rapid miner• R• Python

INFORMATION VISUALIZATION

The best data visualizations are ones that expose something new about the underlying patterns and relationships contained within the data.

Understanding those relationships — and being able to observe them — is key to good decision making

MORE DETAILS: DATA PROCESSING

• Read and write different data types from different data sources

• Process (filter, clean, filter, combine, etc) Data

• Understand Map Reduce concept• Read and Write data from distributed

file system• Work with tools such as R and Hadoop

GROUPING

• Individual preferences 1,2 and 3• We decided, based upon …o Interesto Ambitiono Background

• Not necessarily groups bounded to classes…