Outlier detection algorithms Field of study: Computer Science
Programme code: W4-N2IN19.2021

Module name: Outlier detection algorithms
Module code: W4-IN-N2-20-F-AWOwD
Programme code: W4-N2IN19.2021
Semester:
  • winter semester 2023/2024
  • summer semester 2022/2023
  • winter semester 2022/2023
  • summer semester 2021/2022
Language of instruction: Polish
Form of verification: course work
ECTS credits: 4
Description:
The goal is to introduce the listener to deviation detection algorithms so important in practical applications, e.g. for detecting embezzlement or unusual disease symptoms. Among the issues raised will be both algorithms based on the distance between objects in the analyzed space as well as algorithms derived from cluster analysis that allow identifying objects unlike others and not groupable.
Prerequisites:
(no information given)
Key reading:
Aggarwal, Charu C., „Outlier Analysis”, 2013, Springer Mehrotra, Kishan G., Mohan, Chilukuri, Huang, Huaming, „Anomaly Detection Principles and Algorithms”, 2017, Springer Dhruba Kumar Bhattacharyya, Jugal Kumar Kalita, „Network Anomaly Detection: A Machine Learning Perspective, CRC Press 2014 Hawkins, D. , „Identification of Outliers”, Monographs on Statistics and Applied Probability, Springer, 1980 Tadeusz Morzy , „Eksploracja danych : metody i algorytmy”, Warszawa, 2013,PWN. Larose Daniel „Metody i modele eksploracji danych”. PWN, 2008 Hand David, Manila Heikki, Smyth Padhraic, „Principles of data mining”, MIT Press, 2001.
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]
Is aware of the essence of deviations in the data, which are not errors in the data but real different objects. [M_001]
 K_K01 [1/5]
Has knowledge of the basics of descriptive statistics, including issues to identify deviations in data. [M_002]
 K_W04 [3/5] K_W09 [2/5]
Has knowledge of methods of graphical data representation and detection of deviations in such representations. [M_003]
 K_U09 [3/5]
Has knowledge of selected deviation detection algorithms, including algorithms based on distance and data distribution, as well as algorithms based on data density or local deviations. [M_004]
 K_W02 [2/5] K_W04 [2/5] K_W09 [3/5]
He can choose the right algorithm to detect deviations depending on the type of data being analyzed. [M_005]
 K_U01 [2/5] K_U03 [2/5] K_U08 [2/5] K_U09 [2/5]
Can implement or use ready-made libraries / packages that allow the use of a deviation detection algorithm for a selected data set. [M_006]
 K_U01 [2/5] K_U03 [2/5] K_U08 [2/5] K_U09 [3/5]
Can determine the similarity / distance between two objects in multidimensional space. [M_007]
 K_U01 [2/5] K_U03 [2/5] K_U08 [2/5] K_U09 [3/5]
Type Description Codes of the learning outcomes of the module to which assessment is related
exam (test) [W_001]
Knowledge verification based on the content presented in the lecture. The exam consists of both open and closed theory questions.
 M_001 M_002 M_003 M_004 M_005 M_006 M_007
Projects and reports [W_002]
Developing projects with reports for them within a specified period as a verification of the skills acquired in solving problems.
 M_001 M_005 M_006 M_007
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]
Transferring the content of education in verbal form using audiovisual means and other written teaching aids.
 15
Preparation for the exam.
 15 exam (test) [W_001]
laboratory classes [Z_002]
Detailed preparation of students to solve tasks with an indication of the methodology of the procedure, an indication of the order of performed activities.
 30
Preparation for the laboratory. Student's independent solution of tasks assigned to the laboratory, preparation of reports
 60 Projects and reports [W_002]
Attachments
Module description (PDF)
Information concerning module syllabuses might be changed during studies.
Syllabuses (USOSweb)
Semester Module Language of instruction
(no information given)