omparative benefit of fuzzy logic is that its

omparative Study of ANN and ANFIS for
the Prediction of Employee Turnover in an Organization



Dr. Umang

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now


Department of Manufacturing Process and
Automation Engineering

Netaji Subhas Institute of Technology, New

[email protected]


Mr. Navjot
Singh and Mr. Yashish Swami


Department of Manufacturing Process
and Automation Engineering

Netaji Subhas Institute of Technology, New

 [email protected] &
[email protected]





The purpose of this study is to investigate
the individual employee characteristics and organizational variables that may
lead to employee turnover. Product innovations and corresponding product
variables can be duplicated, but the harmony of a company’s workforce cannot be
replicated, hence they are of utmost importance. Due to this reason, an organization’s
success depends not only on recruiting the new talent but also retaining them.
The study of predicting employee turnover has attempted to explain what factors
make the employees leave and how to prevent the drain of employee talent. If Employee
turnover can be found to be predictable, the identification of at-risk
employees will allow us to focus on their specific needs or concerns in order
to retain them in the workforce. Two classification methods were used to
develop models for predicting employee turnover. Artificial Neural Network
(ANN) and Adaptive neuro-fuzzy inference system (ANFIS).


Keywords: ANN,
ANFIS, Employee turnover, MATLAB





Recently, AI techniques such as Artificial Neural Network
(ANN), Fuzzy Inference System (FIS) and (ANFIS) have been successful in modelling
superiority of human knowledge features. Thus these techniques are of great use
in an environment that is obvious with the absence of a simple and well-defined
mathematical model.


The key
benefit of fuzzy logic is that its knowledge representation is explicit, using simple IF-THEM relations. The Employee Turnover Prediction
cannot be easily described
by artificially explicit knowledge,
because it is
affected by many
unknown parameters. The integration of neural network into the fuzzy logic system, thus the ANFIS, makes it possible to learn from
the prior obtained data sets.

The purposes of this study is to compare the applicability of ANN and ANFIS in
predicting Employee Turnover in an Organization and to identify the most fitted model
to the study area.





The input data used for Employee
turnover prediction are the different
employee characteristics and this data is acquired by Kaggle, an open source dataset platform.

Size of this data set is 15,000 which was normalized
to avoid overfitting of the data. Furthermore, 50 data points has been taken
for prediction.

This graph (Fig.1)
presents the correlations between each variables. The size of the bubbles
reveals the significance of the correlation, while the color presents the
direction (either positive or negative).












Artificial neural network (ANN)


A customized neural network is adopted here. A network first needs
to be trained before using
it for prediction. We have used back-propagation algorithm as the learning algorithm as they are
especially capable of solving prediction problem.


 A total of 15,000 data points were utilized during
training session and 50 data points were used during testing session.
A suitable configuration has to
be chosen for the best performance of the
network. Out of the different configurations
tested, two hidden layer with 50 and 25 hidden neurons
produced the best result. The log sigmoid function was employed as an activation function.
Suitable numbers of epochs have to be assigned to overcome the problem of over fitting
and under fitting of data.







Neuro Fuzzy Inference System (ANFIS)


ANFIS was originally founded by JSR Jang. ANFIS
is a fuzzy system trained on the set of input and output data by an algorithm derived
from the theory of Artificial Neural Networks. This algorithm is a hybrid training
algorithm based on back propagation and the least squares approach.  The ANFIS has advantages such as smoothness
property from the fuzzy principle and adaptability property from the neural networks
training structure.


We have used Subtractive Clustering
algorithm in ANFIS for training the dataset which resolves the problem of
dimensionality and often doesn’t require the optimization from ANFIS commands.













Comparison of ANN and ANFIS models


Results from the
two models are presented in this section to compare the

prediction accuracy.
The same training
testing data sets were used to
train and test both models to extract fair conclusions from the comparison results.







Mean square error (MSE), root
mean square error (RMSE)
were calculated based on the corresponding measured data.





In this paper we
showed the ability of ANFIS and the ANN in predicting the Employee turnover and
potential candidates who are going to leave the firm.

The results showed that the
RMSE, MSE for the training data were 0.088, 0.007 for the ANN model, and 0.160,
0.025 for the ANFIS model. As for unseen data, the RMSE, MSE were 0.8, 0.89 for
the ANN and 0.5, 0.25 for the ANFIS model. The ANFIS model, however, was more
sensitive than the ANN model for the unseen data set and is performing better
for the

We can conclude
that ANN model can fit the output better compared to the ANFIS model for the
unseen data set. But ANFIS is better than ANN in generalization and prediction
of unseen data.