IC091 Optimization of Experiments using Experimental Design and Artificial Neural Networks in Science

Faculty of Science
Autumn 2013
Extent and Intensity
1/0/0. 1 credit(s) (plus extra credits for completion). Type of Completion: z (credit).
Teacher(s)
Prof. E.M. Pena-Mendez (lecturer), prof. RNDr. Josef Havel, DrSc. (deputy)
Guaranteed by
prof. RNDr. Josef Havel, DrSc.
Department of Chemistry – Chemistry Section – Faculty of Science
Supplier department: Department of Chemistry – Chemistry Section – Faculty of Science
Prerequisites (in Czech)
Course INNOLEC will be given in English
Course Enrolment Limitations
The course is offered to students of any study field.
Course objectives
At the end of the course the student should be able: to understand and explain what are artificial neural networks - ANN; to use infromation on ANN for his own work and his data evaluation with ANN; to create INPUT data fro ANN and evaluate them with ANN; to submitt reasoning fro use of ANN in science ; On the base of knowledge achieved to derive new waz fro new data and their evaluation with ANN ; to interpret data from literature about ANN and theirs applications ;to evaluete independentlz his own data;
Syllabus
  • Artificial Neural Networks in Science Prof. Dr. E. M. Pena-Mendez, PhD La Laguna University, La Laguna, Tenerife, Spain Abstract The goal of “Artificial Neural Networks in Science” is to introduce neural networks computing to the students from different branches. Basic theory and mathematical background as well as strategy of use and individual phases: training, verification and prediction will be given but the main aim is via analysing examples from different fields in real life to give the participants also basic practice of ANN use. Therefore, the lectures are combined with practical computing laboratory sessions (informatic room) to gain basic practical experience of ANNs use. Real life applications are related to chemistry, physics, biology and forensic sciences but general ANNs applicability is stressed. The course gives also space to the students to solve their own data and/or problems. Possible participants: students of faculty of science but even those from social sciences. Theory (5 - 6 hrs) 1.- Introduction to neuronal networks computing. History and overview of Artificial Neural Networks (ANN) applications. 2.- ANN basic theory, notation, neuron mode, single-input neuron, transfer functions, multiple-input neuron, network architectures, single and multilayers architectures, ,recurrent networks. 2. SELF ORGANIZING FEATURE MAPS. 3. Classical Statistical analysis vs Neural Networks. 4. OPTIMIZATION in CHEMICAL (BIOLOGICAL) SYSTEMS 4. Application of neural networks for modeling in different fields (Chemistry, Biology, Physics). Computer Practices (15 hours) Lab 1: Supervised Learning 1 Lab 2: Supervised Learning 2 Lab 3: Supervised Learning 3 Lab 4: Unsupervised Learning 1 Consultations Time for personal consulatation and solving participants own data will be assured. Bibliography *Zupan, J., Gasteiger, J., Neural Networks in Chemistry and Drug Design, Wiley VCH, 1999. * J. Havel, E. M. Peña-Méndez, A.Rojas-Hernandez, J-P. Doucet and A. Panaye. Neural Networks for Optimization of high-performance capillary zone electrophoresis methods. A new method using a combination of experimental design and artificial neural networks (ANN). J. Chromatogr. A, 793, 1998, 317-329. * Novotná, H., Vaňhara, J., Tóthová, A., Muráriková, N., Bejdák, P. & Rozkošný, R. Identification and taxonomy of the West Palaearctic species of Tachina Meigen (Tachinidae, Diptera) based on male terminalia and molecular analyses. Entomologica Fennica, 20, 2009, 139-169. * D. Brougham, G. Ivanova, M. Gottschalk, D. M. Collins, A. Eustace, R. O'Connor and J. Havel, Artificial neural networks in metabolomic studies of human lung carcinoma cell lines by in vitro 1H nuclear magnetic resonance of whole cells, J. Biomed. Biotechnol., 2011, ID 158094, 8 pages, doi:10.1155/2011/158094. *Josef Havel; Eladia María Peña Méndez; Alberto Rojas Hernández. ARTIFICIAL NEURAL NETWORKS IN ELECTROPHORESIS. Capillary Electrophoresis and Microchip Capillary Electrophoresis: Principles, Applications and Limitations. Eds. García, C.D. and Carrilho, E.; John Wiley Sons, Inc., Hoboken, NJ, USA. ISBN-10:0470572175, ISBN-13: 978-0470572177 , 2012. *ANN applications in Science. Eladia María Peña Méndez, Marcos Báez Fumero, Josef Havel, Jaromír Vaňhara. Ed. Marcos Báez Fumero and Eladia María Peña Méndez. Depósito Legal, TF 850/2012. GRAFIEXPRESS, S.L., S/C de Tenerife, Spain.
Teaching methods
lectures, discussion, homework, practical drills on computers, individual work with own data
Assessment methods
zápočet
Language of instruction
English
Further comments (probably available only in Czech)
The course is taught only once.
  • Enrolment Statistics (recent)
  • Permalink: https://is.muni.cz/course/sci/autumn2013/IC091