Deep learning with neural networks Field of study: Computer Science
Programme code: W4-S2INA19.2020

Module name: Deep learning with neural networks
Module code: W4-INA-S2-20-F-UGzSN
Programme code: W4-S2INA19.2020
Semester:
  • summer semester 2021/2022
  • winter semester 2021/2022
  • summer semester 2020/2021
Language of instruction: English
Form of verification: course work
ECTS credits: 4
Description:
Basics off deep learning and neural networks. Weights' selection methods and activation functions in neural networks. Classification of neural networks due to their structure. Building deep neural networks. Tuning the parameters of NNs. Reinforcement learning in deep learning.
Prerequisites:
(no information given)
Key reading:
1. Tom Mitchell, Machine Learning, McGraw Hill, 1997. 2. Christopher M. Bishop Pattern Recognition and Machine Learning, Springer, 2007. 3. Paweł Cichosz, Learning Systems, WNT, 2000. 4. S. Russell and P. Norvig, Artificial Intelligence: A Modern Approach, 3rd Ed., Pearson, 2010. 5. D. Poole, A. Mackworth, Artificial Intelligence: Foundations of Computational * Agents, Cambridge University Press, 2010. Weka toolkit, http://www.cs.waikato.ac.nz/ml/weka/. 6. Machine learning with Scikit-Learn and TensorFlow, Aurélien Géron, Helion, 2018. 7. Python: machine learning / Sebastian Raschka, Vahid Mirjalili ; translation: Krzysztof Sawka, Helion, 2019.
Learning outcome of the module Codes of the learning outcomes of the programme to which the learning outcome of the module is related [level of competence: scale 1-5]
The student knows methods and algorithms for learning artificial neural networks. [M_001]
 K_W09 [1/5]
The student can design and implement a deep neural network. [M_002]
 K_W02 [1/5]
The student can design and implement a deep neural network. [M_003]
 K_U01 [1/5]
The student can train a neural network to solve a specific machine learning problem. [M_004]
 K_U02 [1/5]
The student can assess the effectiveness of a trained neural network. [M_005]
 K_U05 [1/5]
The student knows the possibilities of modern neural networks and is aware of the importance of machine learning methods in developing modern IT solutions. [M_006]
 K_K01 [1/5]
Type Description Codes of the learning outcomes of the module to which assessment is related
Test [W_001]
The students solve tasks and answer open-ended questions.
 M_001 M_002
Implementation project [W_002]
The whole group performs multimedia presentation of their project, which is then self-assessed.
 M_003 M_004 M_005
Multimedia presentation [W_003]
The self-assessment of collective work is then evaluated, and the hypotheses are verified.
 M_006
Form of teaching Student's own work Assessment of the learning outcomes
Type Description (including teaching methods) Number of hours Description Number of hours
lecture [Z_001]
The lectures are conducted in the verbal form with the use of content visualisation. They are focused on conceptually difficult material and refer to the addresses of useful websites and e-learning package.
 15
The student internalises the lectures using the existing packages of methods: the script, websites, and the e-learning package.
 30 Test [W_001]
laboratory classes [Z_002]
The classes thoroughly prepare the students for implementing algorithms, emphasising the method and the sequence of operations.
 30
The students study individually for the tests in laboratory classes. They also complete projects – implementations of a chosen system in groups.
 45 Test [W_001] Implementation project [W_002] Multimedia presentation [W_003]
Attachments
Module description (PDF)
Information concerning module syllabuses might be changed during studies.
Syllabuses (USOSweb)
Semester Module Language of instruction
(no information given)