In this course, we are going to follow some nice O'Reilly data science manual and, line by line, learn about meaning of terms like "feature", "multi-class classification", "training" and "cross validation" and, while doing so, acquire all necessary prerequisities of "the most sexy job of 22nd century".

We start this Friday (24th April) at 10:00 am

Session 0 :: Warm-up

Order of the day


  • 10:00 - 10:15 :: Digitales Ankommen

  • Introduction

  • Course overview

  • Machine-learning intro

  • Jupyter intro

  • Warm-up exercise (in break-out rooms ??)

  • Discussion

  • Warm-up exercise (in break-out rooms ??)

  • Discussion

Who am I ? (Daniel)

  • founder of oldest Slovak digital community kyberia.sk
  • Bc. in humanities (Charles University in Prague) and Bc. in linguistics (Universite de Nice Sophia-Antipolis)
  • MSc. in cognitive sciences (Ecole Pratique des Hautes Etudes, Paris)
  • PhDs. in psychology (Universite Paris 8) and cybernetics (Slovak University of Technology)
  • ex IT-Admin of UdK's Medienhaus
  • Digital Education juniorprofessor (Einstein Center Digital Future / UdK)

Who am I ? (Paul)

  • Artist and Programmer
  • Wissenschaftlicher Mitarbeiter von Daniel
  • BA in Visuelle Kommunikation
  • plsdlr.net

Who am I ? (Nik)

  • Artist and Electrical Engineer
  • Studentische Hilfskraft von Daniel
  • BA in Electrical Engineering / KuM Absolvent(Almost)

M.O.D.E.A #4 : (Machine) learning and data (science)

In this course, we are going to follow some nice O'Reilly data science manual and, line by line, learn about meaning of terms like "feature", "multi-class classification", "training" and "cross validation" and, while doing so, acquire all necessary prerequisities of "the most sexy job of 22nd century".

We start this Friday (24th April) at 10:00 am

Session 0 :: Warm-up

Session 1 :: 15th May :: Features 1

Session 2 :: 12th June :: Classifiers







12.6.2020 / AE500612 :: Neighbors & Classifiers

Session 3 :: 26th June :: Evaluation & Tradeoffs

Session 4 :: 19th June :: Tradeoff

Session 5 :: 3rd July :: Integration & Deployment

What is this course NOT about ?

it is not about neural networks *


it is not about "regression"


 


* well, it will be also about neural networks, but just a little bit ;)

What is this course about ?

about learning


about classification


about main machine learning concepts: true positive / true negative / false positive / false negative / feature, feature extraction, classifier, accuracy etc.

Session 1 :: 15th May :: Features 1

Core principle

Success of Your machine learning approach depends in great measure from quality of Your features.

Session 2 :: 12th June :: Classifiers







12.6.2020 / AE500612 :: Neighbors & Classifiers

Teasers

Teaser 0 :: Master in Design & Computation

TU / UdK Master Program
https://www.design-computation.berlin/
apply until 31.7.2020

Teaser 1 :: Umfrage zur digitalen Semester

https://survey.wi.uni-potsdam.de/index.php/152432?newtest=Y&lang=de

Teaser 2 :: Bäume der Erkenntnis

https://www.arte.tv/de/videos/RC-015184/baeume-der-erkenntnis/

Your projects

#clover AI

Johannes

Fred

scene change detection

finding differences between two images

uses OpenCV

method 1 :: based on subtraction of consecutive frames, RGB values of result of subtraction are thresholded and yield a binary image; finally, one counts the amount of white pixels

method 2 :: uses histogram distributions

Classifiers

Support Vector Machines

An SVM model is a representation of the examples as points in space, mapped so that the examples of the separate categories are divided by a clear gap that is as wide as possible. New examples are then mapped into that same space and predicted to belong to a category based on the side of the gap on which they fall.

Logistic Regression

k-Nearest Neighbors

Next sessions

19.6. 12:00 - 13:30 Digital
26.6. 12:00 - 13:30 Digital
10.7. 10:00 - ?? Physical (Kleistpark)

Session 3 :: 26th June :: Evaluation & Tradeoffs

Session 4 :: 19th June :: Tradeoff

Session 5 :: 3rd July :: Integration & Deployment