Systems analysis and decision support methods in Computer Science 
(the exercise class, summer 2019)

 

 

Course information

Office hours:

  • Friday, 18:15-20:35, 121 C3 (after e-mail notification) 
  • Wednesday, 18:15-20:35, 121 C3 (after e-mail notification) 

Credit rules: [link]

Recommended literature:

  • Murphy, Kevin P. Machine learning: a probabilistic perspective. MIT press, 2012.
  • Bishop, Ch. Pattern Recognition and Machine Learning. Springer, 2006. 

 

Additional materials:

  • Machine learning course [link

The link to my teaching calendar (summer semester, all groups) [link]

 

Results: [link] (link is inactive when the file is modified)

 

Course schedule

Calendar with marked deadlines for each exercise list: [link] (updated, 31.03.2019)

 

Class I: Organizational issues

Prerequisites: -

Class schedule: 

  • Safety rules, course credit rules, organizational issues
  • Introduction to machine learning

Additional materials: -

 

Class II: Gradients and linear algebra

Prerequisites:

  • Completed assignment 1 [link]
  • Submitted declaration [link]

Class schedule: 

  • Presentation of mathematical exercises (assignment 1)
  • Theoretical introduction to assignment 2

Additional materials:

  • The program of "Linear Algebra" and "Mathematical Analysis" courses
  • Basic mathematical terms: gradient [link], norm [link],
    inner product [link], matrix multiplication [link]

 

Class III: ML and MAP estimation

Prerequisites:

  • Completed assignment 2 [link]
  • Submitted declaration [link]

Class schedule: 

  • Presentation of mathematical exercises (assignment 2)
  • Theoretical introduction to assignment 3

Additional materials:

  • Basic mathematical terms:
    maximum Likelihood estimation (ML) [link],
    maximum a posteriori estimation (MAP) [link],
    Bernoulli distribution [link],
    categorical distribution [link],

    Multivariate normal distribution [link]
  • Murphy:
    from chapter 2: 2.3.1, 2.3.2, 2.4.5, 2.5.2,
    from chapter 3: 3.2.1-3.2.3, 3.3.1-3.3.3

 

Class IV: Regression (part I)

Prerequisites:

  • Completed assignment 3 [link]
  • Submitted declaration [link]

Class schedule: 

  • Presentation of mathematical exercises (assignment 3)
  • Theoretical introduction to assignment 4

Additional materials:

  • Murphy:
    from chapter 7: 7.1, 7.2, 7.3

Class V: Regression (part II)

Prerequisites:

  • Completed assignment 4 [link]
  • Submitted declaration [link]

Class schedule: 

  • Presentation of mathematical exercises (assignment 4)
  • Theoretical introduction to assignment 5

Additional materials:

  • Murphy:
    from chapter 7: 7.1, 7.2, 7.3

 

Class VI: Classification (part I)

Prerequisites:

  • Completed assignment 5 [link]
  • Submitted declaration [link]

Class schedule: 

  • Presentation of mathematical exercises (assignment 5)
  • Theoretical introduction to assignment 6

Additional materials:

  • Murphy, 4.1, 4.2

 

Class VII: Classification (part II)

Prerequisites:

  • Completed assignment 6 [link]
  • Submitted declaration [link]

Class schedule: 

  • Presentation of mathematical exercises (assignment 6)

Additional materials:

  • Murphy, 8.1-8.3

 

SELECTED PROJECTS

 

Deep convolutional neural network
for an automated signal selection in protein NMR spectra

 

Chemical shift-based identification
of oligosaccharide topology

Image analysis for confocal microscopy in the context of 3D liver tissue architecture studies

Commercial projects

X 2014 - ...

Read more

II 2016 - VI 2018

Read More

VII 2013 - VII 2014

Read More

  

Read More

 

 

RECENT WORK

 

2018/6/6

Our manuscript "Chemical shift-based identification of monosaccharide spin-systems with NMR spectroscopy to complement untargeted glycomics" has been accepted in Bioinformatics. 

Open access article is available here. The algorithm proposed in the paper is accessible by the web page glyconmrsearch.nmrhub.eu

 

2018/5/28

Our manuscript "Towards fully automated protein structure elucidation with NMR spectroscopy" has been accepted at BOOM workshop (ICML 2018).

 

 

2018/3/4

Our manuscript "NMRNet: A deep learning approach to automated peak picking of protein NMR spectra" has been accepted in Bioinformatics.

 

CONTACT DETAILS

E-mail:
Building C-3, room 121
Department of Computer Science
Wrocław University of Technology
Wybrzeże Stanisława Wyspiańskiego 27
50-370 Wrocław, Poland