Volume 3 : Issue 2, June 2014

Table of Contents, 30 June 2014

Research Title/
Graphical Abstract

Article Information/
Abstract

Download

PII: S238309801400003-3

Implementation of Statistical Quality Control in Dairy Factories Using Univariate and Multivariate Control Charts


Original Research, C3

Ahmadvand M, Hosseini Nasab H, Golmohammadi A-M.

Sci. J. Mech. Ind. Eng., 3(2): 18-26, 2014

ABSTRACT: Statistical quality control, the use of statistical techniques is in all stages of production, processing and marketing services, so we can have quality products in accordance with the desired characteristics of the consumers and make sure customer satisfaction. One of the main tools of statistical process control is the use of control charts. When control charts are most commonly exploited the qualitative characteristics of the study was the univariate and comply normal distribution. If the simultaneously study of qualitative characteristics to be important, parameters monitoring methods for multivariate process is prepared, but this need to know more statistical methods. Many analysts prefer the simplicity of calculations of independence hypothesis quality characteristics considered and their issues in the univariate statistical environment are examined. In this paper, a univariate control charts for technical specifications dairy prepared then Hotelling multivariate control charts for variables prepared and the results are compared with each other.
Key Words:
Statistical Quality Control, Univariate Control Charts, Multivariate Control Charts, Hotelling

pdfS_icon.

PII: S238309801400004-3

Prediction using a hybrid particle swarm optimization and GARCH model


Original Research, C4

Narimani R, Narimani M.

Sci. J. Mech. Ind. Eng., 3(2): 27-32, 2014

ABSTRACT: This paper addresses one of the most important data mining tasks, forecasting, where the objective is to predict any time series. In order to achieve the best performance. At the first stage, we find the best GARCH model by looping through AR and MA term. At the next stage, we propose PSO-GARCH model that optimize the coefficient of the model. Particle swarm optimization is one of the best optimization techniques that is suitable for continues variable, thus by using PSO we can find the best coefficient of GARCH model obtained from first stage. The fitness function gets the weights from PSO particle and optimizes the weights by using Marquardt algorithm. In our experiment, two dataset were used: The daily observations of Brent oil price and the exchange rate of US Dollar to Euro (USD2EUR). The achieving results show that the performance of the hybrid method is so promising.
Key words:
PSO, GARCH, PSO-GARCH prediction

pdfS_icon