|
Kategorie: Diplomové, bakalářské práce |
Tento dokument chci!
Práce popisuje základy principu funkčnosti neuronů a vytvoření umělých neuronových sítí. Je zde důkladně popsána struktura a funkce neuronů a ukázán nejpoužívanější algoritmus pro učení neuronů. Základy fuzzy logiky, včetně jejich výhod a nevýhod, jsou rovněž prezentovány. Detailněji je popsán algoritmus zpětného šíření chyb a adaptivní neuro-fuzzy inferenční systém. Tyto techniky poskytují efektivní způsoby učení neuronových sítí.
Each term represents fuzzy set. stage definition the rules according the firing strengths the inputs
3. stage conversion numerical input value fuzzy value fuzzyfication
2. stage retransformation the resultant fuzzy values into numerical values -
defuzzyfication
Main advantages the fuzzy systems:
ability represent uncertainties the human knowledge with linguistic
variables
easy interpretation the results
easy expansion the base knowledge addition new rules
robustness relation the possible disorders the system
Main disadvantages are:
unable universalize, only answers what written its rule base
topological changes the system would demand alternation the rule base
definition the inference logical rules needs expert
4. The fuzzy interference
mechanism consists three stages:
1. Its development was
motivated the need for conceptual framework, which can help addressing the
issue uncertainty and lexical imprecision. Neural net technology can used learn system
. FUZZY SYSTEMS
Fuzzy logic was first developed 1965 Lotfi Zadeh.17
4. The terms the input space (typically 5-7
for each linguistic variable) compose the fuzzy partition [1]. Some significant characteristics the fuzzy logic are:
fuzzy logic, exact reasoning viewed limiting case approximate
reasoning [6]
fuzzy logic, everything matter degree [6]
fuzzy logic, knowledge interpreted collection elastic or, equivalently,
fuzzy constrain collection variables [6]
Inference viewed process propagation elastic constraints [6]
Any logical system can fuzzified [6]
The function such systems can described set fuzzy rules, like ‘if-
then’ (premise-consequent). With the help fuzzy logic the
uncertainties human cognitive processes like thinking and reasoning can be
expressed mathematically. Fuzzy logic uses graded statements rather than ones that
are strictly true false. provides an
approximate but effective means describing behavior systems that are too
complex, ill-defined not easily analyzed mathematically.1 Fuzzy Neural Networks
A marriage between fuzzy logic and neural networks can attenuate the
problems these technologies. If-then rules use linguistics variables with symbolic
terms