SBS

Simulation of Biological Systems (E374013)

Artificial live

The "life game" simulation
- Cellular automata (_)
Swarm intelligence (study of nature, not the numerical optimization method)
Artificial ants (vant-s, models of ants' behaviour, and the numerical optimization method of this name; yt)
(other - study of the state of arts)

Modeling of process

differential equation for dynamic description (possibly: how to solve using Matlab+Simulink)
- population simulation - single population, dual (hare-wolf), three populations (+vegetation grow)
compartment method for drug in body distribution
SIR model
deterministic chaos, Lorentz attractor
catastrophic theory, examples of use

Identification

using known model, guessing parameters (gradient descent)
using known model, guessing parameters (using swarm/genetic algorithms)
using neural network (can an artificial neural network work as a model?)
data analysis (correlation, autocorrelation, Fourier transform)

Artificial/biological neuron

biological neuron description
artificial neuron network, perceptron, MLP, backpropagation, teaching coefficient
other types (radial based, Hopfield network)
- homeostatic neural network, Růžek (CZ), Růžek, Brandejský

AI: Genetic algorithm

Proposed method of classification

Each student will prepare monothematic presentation, in a range of 20 - 90 minutes, minimum 45 for full classification, including examples, and a pdf version of notes for supporting study to exam.

Student will present it on classes.

Other students will declare recommended number of points for classification.

50 points from the classification (how sucessfully the student classify the other students)
- 20 points - given classification,
- 30 points - asking questions (other students can be prepared for the topic, for example on the base of the Wikipedia links, presented in this page).
50 points from the classification given by other students. This classification could be normalized.

This year schedule (limited validity)

Dr. Vrana