On Wednesday, July 1 2020, ISTTS held an open online KSP
for internal and external. This time KSP raised the topic "Fuzzy Decision in Fuzzy
Environment "and was delivered by an experienced ISTTS lecturer, Mr
Judi Prajetno Sugiono, Ir., M.M.
Mr. Sugiono started KSP by explaining the comparison of fuzzy logic
with classical logic. Classical logic uses a number base, emphasizing accuracy
(precision), cut the level of existence of random elements so that the boundary of the problem is clear,
requires a long design time and expensive design costs. Whereas logic
fuzzy uses a base of words (linguistics) so that it is more significant and contains
imprecision. This logic design time is faster and its design costs
also needed cheap. In addition, fuzzy logic also considers random elements
so that the problem boundaries are not explicitly cut.
"The fuzzy logic is not clear but the linguistic expression that is produced is logic
fuzzy is more significant than classical logic which gives deep expression
precise numeric form, "concludes Mr. Sugiono about the difference in fuzzy logic with
classical logic.
"There are many things we can get from scientific Fuzzy Logic," said Mr. Sugiono.
He also explained the nature of fuzzy logic, among others: conceptually easy
understood, flexible, tolerant of inaccurate data, can model functions
nonlinear complexity of change, can be built on the experience of para
expert, can be mixed with conventional control techniques and based on natural language.
"Conceptually, [fuzzy logic] is easy to understand. Like the temperature of tea in a glass: cold, hot
and warm. Fuzzy is also flexible in the sense that it can be done in many ways and more
easy to understand, "he added.
Mr. Sugiono explained that with fuzzy logic, random elements could still be accepted,
different from classical logic which discards random elements. "There are elements
indecisiveness that is not thrown away and sometimes it helps people to
solve the problem, "explained Mr. Sugiono about the reason why using logic
fuzzy.
Fuzzy Inference System consists of several processes including: Fuzzification (process
convert crisp inputs / numbers into fuzzy input / words), Fuzzy Rule
(IF statement ... THEN ... which is used to express the relationship between fuzzy
input), Inference Mechanism (evaluation of the fuzzy rule used to infer
output / to infer an output) as well as Defuzzification (the process for changing a conclusion
achieved by the inference mechanism being the input of the next process).
The range of the use of fuzzy logic a lot like: Fuzzy Logic Controller,
Fuzzy Expert System, Fuzzy Data Clustering, Fuzzy Decision Making, Neuro Fuzzy Inference
System and Fuzzy Image Processing. "Because of the fuzzy nature whose operations are isolated from the input
or the actual output, the fuzzy logic 'seems to be pasted' on
existing algorithms, "said Mr. Sugiono.
After going through various processes, fuzzy logic can be used to help
decision making process. Fuzzy decision making is a collection of techniques
aims to choose the best alternative in the situation of inaccurate data availability, no
complete and unclear. One example is helping investment decision making.