infrastructure is very supportive in the work to
facilitate and expedite work, especially work space.
The results of this study prove that the
completeness of workspace infrastructure facilities
affects employee job satisfaction. Employees who
state that the work space is complete and in good
condition tend to feel satisfied at work. The
management realizes that employee participation
must be considered because employees as human
resources play the most important and potential role
for hospital success. Completeness of infrastructure
related to work space in the form of area, cleanliness,
comfort and layout of photos or paintings.
With the completeness of the furniture provided,
the employee will feel satisfied at work. Biasana
satisfaction refers to the pleasure and love of his
work. In addition, employees also demonstrate
discipline by complying with predetermined rules and
demonstrate work performance both individually and
in groups.
3.2 Furniture Effects
It was found that there is influence of furniture on
employee job satisfaction in hospitals. Staff who
stated that the hospital furniture was good, had a 8.6
times higher chance of being satisfied compared to
those who did not.
The results of this study prove that furniture has a
significant effect on employee job satisfaction.
Employees who state that furniture is complete and in
good condition tend to feel satisfied compared to
incomplete ones. The completeness and availability
of furniture in the employee's office is related to the
arrangement of the location of furniture which is well
organized and neat, ergonomic chairs and tables,
cupboards to store adequate files, furniture conditions
that have long been replaced. All of that is to support
employee work and increase job satisfaction.
3.3 Office Supplies Effects
Based on the results of the study showed that there is
an influence of office equipment on job satisfaction
of hospital employees. Employees who state good
hospital office equipment have a 7.9 percent higher
chance of being satisfied compared to less good ones.
It was found that the completeness of the office
affects employee job satisfaction. Employees who
state that complete office equipment tends to feel
satisfied at work. The work environment can support
work improvement, the office was more attentive and
is able to provide office equipment to support the
convenience of employees at work. The availability
of office equipment needed by employees will
improve and accelerate work to work more
effectively and efficiently.
3.4 Smart Model with ANFIS
Data processing starts from entering data, designing
input and output forms, trying out training data, and
continuing with testing the training data. Next, the
model adjusted the range of values and shapes of the
input pyramid. 40% of the total data is used as
training data and the rest is used as test data. Figure 2
(a) shows the selected training data. Inputs, outputs,
and rules on anfis are set using Grid Partition.
Member Functions (MF) serves as a reference for the
distance value from the input that will categorize the
level of assessment of the respondent. In this
experiment, 3 MF to provide 3 categories of
assessment based on the value of the respondents
were used. For the output itself, a constant value is
used to increase the accuracy of the data to be
processed. In the end some rules that are related to
each other will be formed by themselves.
The accuracy of the modeling results is improved
by using a hybrid method. The use of hyrid is a
combination of backpropagation and least-squares
regression which aims to adjust the FIS parameters.
Error tolerance is given a value of 0 which indicates
that training will stop when the amount of training is
reached. The value of the epochs is given 10 which
indicates the data will be given 10 repetitions in order
to get maximum accuracy. Figure 2 (b) shows the
results of testing toward the training data. The results
on the data (blue dot) show that the test and training
data are coincidental (red dot) which indicate the high
level of modeling accuracy. It shows the data that we
present with our target of fulfilling what we want.
Figure 3 is a form of rule structure that has been
composed of various numbers in the data. The data
consists of 7 inputs in the form of Building, work
space, lighting, furniture, communication tools,
office supplies, and air fresheners and 1 output in the
form of employee job satisfaction. The rules formed
are interconnected with all input and possibilities
formed to produce output.
As explained in the previous Grid Partition, the
model will be formed as shown in Figure 3. Consists
of three pyramids that indicate the level of
respondents' assessment to be processed. Consists of
less, medium, and good, each of which represents the
category for the number obtained from the
respondent's satisfaction value. The distance value
used is 1 - 5. The use of this trimf model is very
suitable for this data because the accuracy obtained is
quite good.
Designing a Smart Prediction Model for Influence of the Infrastructure Completeness on Work Satisfaction