Programme & Teachers

Photos by Ferran Larriba Pérez


Sunday 02/09/2018

Social MeetUp: For early arrived students, we are planning to have a Tapas route (each one pays his own) at the Borne neighborhood. The meeting point is: "Palau de la Generalitat de Catalunya" (https://goo.gl/maps/f34FDdMx3H72) at 18.00h. Feel free to join us!

Monday 03/09/2018

Topic: Feature Selection.

Professor: Dr. Andrzej Janusz

Affiliation: University of Warsaw, Institute of Computer Science.

Syllabus:

9.00h Training

  1. Feature selection – an introduction
    1. Feature Selection Problem – what is it?
    2. Do we still need FS (and why we do)?
    3. Some examples of practical problems where FS is a must
  2. Categorization of FS methods
    1. Filtering approaches with a discussion of their pros and cons
    2. Wrapper approaches with a discussion of their pros and cons
    3. Embedded methods
  3. Examples of filtering approaches
    1. The contextuality - univariate/multivariate methods
      • Correlation filters
      • Statistical tests
      • Entropy/gini/discernibility-based filters
      • Relief/reliefed algorithm
    2. Tuning feature scores into feature subsets – practical issues
      • K-best approach
      • Selecting optimal ‘k’ – permutation tests
    3. Maximal relevance – minimal redundancy rule (MRMR)
    4. Rough sets and decision reducts

10.30h Coffee break

11.00h Training

  1. Examples of the wrapper approach and feature subset search strategies
    1. The greedy heuristic
    2. Recursive feature elimination
    3. Examples of different heuristics
  2. Examples of embedded methods
    1. LASSO regression
    2. Role of the L1 regularization
  3. Examples of other notable FS methods which are difficult to classify
    1. RF-based feature rankings
    2. Ensembles of feature subsets
    3. Feature clustering and its uses
    4. The notion of feature interchangeability

13.00h Lunch breaks

14.15h Training

  1. R language and its most popular FS libraries
  2. Methodology of measuring performance of FS methods
  3. Golub’s experiment

15.45h Coffee break

16.00 Training

  1. Extraction of features from sensory data obtained from coal mines
  2. Performance tests of several FS methods
  3. Investigation of feature interchangeability

17.00 End of the day

Tuesday 04/09/2018 - Wednesday 05/09/2018

Topic: Microsoft Azure for Life Sciences.

Professors: Dr. Dimitrios Vlachakis

Teacher Assistance: Eleni Papakonstantinou

Teacher Assistance: Louis Papageorgiou

Affiliation: Genetics Laboratory, Agricultural University of Athens.

Syllabus:

  1. Introduction to MS Azure; applications and future scope
  2. Hands on #1: Data management and Analytics in MS Azure
  3. Hands on #2: Speeding up research in the MS Azure cloud-omics ecosystem

Pot Luck Wednesday 05/09/2018: We encourage you to bring snacks, food, beverages from your country to share them after the classes finished!

Thursday 06/09/2018 - Friday 07/09/2018

Topic: Deep Learning.

Professors: Dr. Tarry Singh

Teacher Assistance: Raik Otto

Affiliation:Founder and CEO, deepkapha.ai

Syllabus:

Thursday 06/09/2018

9am - 12 pm : Title — Deep Learning Introduction (Computer Vision and CNNs) | Mode — Theory

12 - 5 pm : Title — Putting theory of Computer Vision into practice | Hands-on coding

Friday 07/09/2018

9am - 11 pm : Title — Use of Deep Learning in Healthcare / Bioinformatics | Mode — Theory

11am - 5 pm : Title — Detecting skin Cancer or Breatmass Cancer with code | Mode — Hands-on coding