Module also offered within study programmes:
General information:
Name:
Data mining and machine learning in civil engineering problems
Course of study:
2019/2020
Code:
GIKS-2-402-IS-n
Faculty of:
Mining and Geoengineering
Study level:
Second-cycle studies
Specialty:
Environmental Installations
Field of study:
Environmental Engineering
Semester:
4
Profile of education:
Academic (A)
Lecture language:
Polish
Form and type of study:
Part-time studies
Course homepage:
 
Responsible teacher:
dr hab. inż, prof. AGH Rusek Janusz (rusek@agh.edu.pl)
Module summary

Problems of classification and regression as well as dimensionality reduction in data sets on building
objects and operational impacts using Data-Mining and Machine Learning methods.

Description of learning outcomes for module
MLO code Student after module completion has the knowledge/ knows how to/is able to Connections with FLO Method of learning outcomes verification (form of completion)
Social competence: is able to
M_K001 work in a team while doing design exercises and exchange information. Project
Skills: he can
M_U001 - create regression and classification models based on available data using the Python programming environment and the Matlab platform, - use models for individual simulation purposes and integration with other parts of the code, - assess the significance of the impact of individual input variables on the final neural network or SVM system. Project
M_U002 - form classification and regression problems based on data, - separate the training, validation and test set for artificial neural network and SVM models, - write optimization algorithms used to build neural networks. Test
Knowledge: he knows and understands
M_W001 - formulation of the objective function used in the process of learning artificial neural networks (ANN) and related systems, - regularization and stabilizing methods of the learning process, - methods of gradient optimization and derivative-free optimization on the example of Pattern Search and genetic algorithms and related methods (PSO etc.), - construction of the objective function for the Support Vector Machine (SVM) method in the classification and regression problem, - methods for assessing the effectiveness of learning in terms of generalization. Test
M_W002 - methods for assessing the quality of data for further processing, - Data-Mining methods used to reduce the dimensionality of data and extract important features (feature selection), - methods of preliminary analysis of the linear relationship between variables. Basic concepts from the area of classification and regression models. Test
Number of hours for each form of classes:
Sum (hours)
Lecture
Audit. classes
Lab. classes
Project classes
Conv. seminar
Seminar classes
Pract. classes
Zaj. terenowe
Zaj. warsztatowe
Prace kontr. przejść.
Lektorat
15 9 0 0 6 0 0 0 0 0 0 0
FLO matrix in relation to forms of classes
MLO code Student after module completion has the knowledge/ knows how to/is able to Form of classes
Lecture
Audit. classes
Lab. classes
Project classes
Conv. seminar
Seminar classes
Pract. classes
Zaj. terenowe
Zaj. warsztatowe
Prace kontr. przejść.
Lektorat
Social competence
M_K001 work in a team while doing design exercises and exchange information. - - - + - - - - - - -
Skills
M_U001 - create regression and classification models based on available data using the Python programming environment and the Matlab platform, - use models for individual simulation purposes and integration with other parts of the code, - assess the significance of the impact of individual input variables on the final neural network or SVM system. - - - + - - - - - - -
M_U002 - form classification and regression problems based on data, - separate the training, validation and test set for artificial neural network and SVM models, - write optimization algorithms used to build neural networks. - - - + - - - - - - -
Knowledge
M_W001 - formulation of the objective function used in the process of learning artificial neural networks (ANN) and related systems, - regularization and stabilizing methods of the learning process, - methods of gradient optimization and derivative-free optimization on the example of Pattern Search and genetic algorithms and related methods (PSO etc.), - construction of the objective function for the Support Vector Machine (SVM) method in the classification and regression problem, - methods for assessing the effectiveness of learning in terms of generalization. + - - + - - - - - - -
M_W002 - methods for assessing the quality of data for further processing, - Data-Mining methods used to reduce the dimensionality of data and extract important features (feature selection), - methods of preliminary analysis of the linear relationship between variables. Basic concepts from the area of classification and regression models. + - - + - - - - - - -
Student workload (ECTS credits balance)
Student activity form Student workload
Summary student workload 78 h
Module ECTS credits 3 ECTS
Udział w zajęciach dydaktycznych/praktyka 15 h
Preparation for classes 10 h
przygotowanie projektu, prezentacji, pracy pisemnej, sprawozdania 20 h
Realization of independently performed tasks 30 h
Examination or Final test 2 h
Contact hours 1 h
Module content
Lectures (9h):

The course will discuss methods in the field of Data Mining and Machine Learning. The mathematical foundations of these methods and examples of their application will be presented. The problems of classification and regression in a linear and nonlinear approach will be discussed. The methods of reducing the size of the data originally collected will be presented. Examples will be presented of the use of these methods in the area of information about civil engineering construction objects.

Project classes (6h):

Execution of design exercises within which, for an available set of data on the technical characteristics of construction objects and operational impacts, classification and regression problems will be formulated. Methods of artificial neural networks and the SVM method will be used for this. Finished models will be tested for correctness and generalization. Different optimization methods for teaching artificial neural networks will be tested. The SVM method will analyze the selection of parameters resulting from the original form of the objective function.

Additional information
Teaching methods and techniques:
  • Lectures: Treści prezentowane na wykładzie są przekazywane w formie prezentacji multimedialnej w połączeniu z klasycznym wykładem tablicowym wzbogaconymi o pokazy odnoszące się do prezentowanych zagadnień.
  • Project classes: Studenci wykonują zadany projekt samodzielnie, bez większej ingerencji prowadzącego. Ma to wykształcić poczucie odpowiedzialności za pracę w grupie oraz odpowiedzialności za podejmowane decyzje.
Warunki i sposób zaliczenia poszczególnych form zajęć, w tym zasady zaliczeń poprawkowych, a także warunki dopuszczenia do egzaminu:

attendance at classes: obligatory (2 unjustified absences)
attendance at lectures: indicated (optional)

Participation rules in classes:
  • Lectures:
    – Attendance is mandatory: No
    – Participation rules in classes: Studenci uczestniczą w zajęciach poznając kolejne treści nauczania zgodnie z syllabusem przedmiotu. Studenci winni na bieżąco zadawać pytania i wyjaśniać wątpliwości. Rejestracja audiowizualna wykładu wymaga zgody prowadzącego.
  • Project classes:
    – Attendance is mandatory: Yes
    – Participation rules in classes: Studenci wykonują prace praktyczne mające na celu uzyskanie kompetencji zakładanych przez syllabus. Ocenie podlega sposób wykonania projektu oraz efekt końcowy.
Method of calculating the final grade:

Degree of the test (lecture) X 0.5 + Degree of the project X 0.25 + Degree of the test (classes) X 0.25

Sposób i tryb wyrównywania zaległości powstałych wskutek nieobecności studenta na zajęciach:

in the student’s own scope

Prerequisites and additional requirements:

- basic knowledge of mathematical analysis, algebra and probabilistics,
- ability to act on vectors and matrices.

Recommended literature and teaching resources:

OSOWSKI, Stanisław. Metody i narzędzia eksploracji danych. Legionowo: Wydawnictwo BTC, 2013.
OSOWSKI, S. Sieci neuronowe do przetwarzania informacji, Oficyna Wydawnicza Pol. Warszawskiej, Warszawa, 2006.
ŁĘSKI, Jacek. Systemy neuronowo-rozmyte. Wydawnictwa Naukowo-Techniczne, 2008.
CICHOSZ, Paweł. Systemy uczące się. Wydawnictwa Naukowo-Techniczne, 2000.

Scientific publications of module course instructors related to the topic of the module:

RUSEK, Janusz. Modelowanie stopnia zużycia technicznego budynków na terenach górniczych z
wykorzystaniem wybranych metod sztucznej inteligencji. Wydawnictwa AGH, 2013.
RUSEK, Janusz. Creating a model of technical wear of building in mining area, with utilization of
regressive SVM approach. Archives of Mining Sciences, 2009, 54.3: 455–466.
FIREK, K.; RUSEK, Janusz; WODYŃSKI, Aleksander. Wybrane metody eksploracji danych i uczenia
maszynowego w analizie stanu uszkodzeń oraz zużycia technicznego zabudowy terenów górniczych.
Przegląd Górniczy, 2016, 72.1: 50—55.

Additional information:

-