Abstract: Data mining uses a variety of techniques each of which is useful for some particular task. It is important to have a deep understanding of each technique and be able to perform sophisticated analysis. In this article we describe a tool built to simulate a variation of the Kohonen network to perform unsupervised clustering and support the entire data mining process up to results visualization. A graphical representation helps the user to find out a strategy to optmize classification by adding, moving or delete a neuron in order to change the number of classes. The tool is also able to automatically suggest a strategy for number of classes optimization.The tool is used to classify macroeconomic data that report the most developed countries? import and export. It is possible to classify the countries based on their economic behaviour and use an ad hoc tool to characterize the commercial behaviour of a country in a selected class from the analysis of positive and negative features that contribute to classes formation.
Abstract: In order to define a new model of Tunisian foot
sizes and for building the most comfortable shoes, Tunisian
industrialists must be able to offer for their customers products able
to put on and adjust the majority of the target population concerned.
Moreover, the use of models of shoes, mainly from others
country, causes a mismatch between the foot and comfort of the
Tunisian shoes.
But every foot is unique; these models become uncomfortable for
the Tunisian foot. We have a set of measures produced from a
3D scan of the feet of a diverse population (women, men ...) and we
try to analyze this data to define a model of foot specific to the
Tunisian footwear design.
In this paper we propose tow new approaches to modeling a new
foot sizes model. We used, indeed, the neural networks, and specially
the Kohonen network.
Next, we combine neural networks with the concept of half-foot
size to improve the models already found. Finally, it was necessary to
compare the results obtained by applying each approach and we
decide what-s the best approach that give us the most model of foot
improving more comfortable shoes.
Abstract: Partial discharge (PD) detection is an important
method to evaluate the insulation condition of metal-clad apparatus.
Non-intrusive sensors which are easy to install and have no
interruptions on operation are preferred in onsite PD detection.
However, it often lacks of accuracy due to the interferences in PD
signals. In this paper a novel PD extraction method that uses frequency
analysis and entropy based time-frequency (TF) analysis is introduced.
The repetitive pulses from convertor are first removed via frequency
analysis. Then, the relative entropy and relative peak-frequency of
each pulse (i.e. time-indexed vector TF spectrum) are calculated and
all pulses with similar parameters are grouped. According to the
characteristics of non-intrusive sensor and the frequency distribution
of PDs, the pulses of PD and interferences are separated. Finally the
PD signal and interferences are recovered via inverse TF transform.
The de-noised result of noisy PD data demonstrates that the
combination of frequency and time-frequency techniques can
discriminate PDs from interferences with various frequency
distributions.
Abstract: The production of activated carbon from low or zero cost of agricultural by-products or wastes has received great attention from academics and practitioners due to its economic and environmental benefits. In the production of bamboo furniture, a significant amount of bamboo waste is inevitably generated. Therefore, this research aimed to prepare activated carbons from bamboo furniture waste by chemical (KOH) activation and determine their properties and adsorption capacities for water treatment. The influence of carbonization time on the properties and adsorption capacities of activated carbons was also investigated. The finding showed that the bamboo-derived activated carbons had microporous characteristics. They exhibited high tendency for the reduction of impurities present in effluent water. Their adsorption capacities were comparable to the adsorption capacity of a commercial activated carbon regarding to the reduction in COD, TDS and turbidity of the effluent water.
Abstract: The reservoir of Kowsar dam supply water for
different usages such as aquaculture farms , drinking, agricultural and
industrial usages for some provinces in south of Iran. The Kowsar
dam is located next to the city of Dehdashat in Kohgiluye and
Boyerahmad province in southern Iran. There are some towns and
villages on the Kowsar dam watersheds, which Dehdasht and Choram
are the most important and populated twons in this area, which can to
be sources of pollution for water reservoir of the Kowsar dam . This
study was done to determine of water pollution of the Kowsar dam
reservoir which is one of the most important water resources of
Kohkiloye and Boyerahmad and Bushehr provinces in south-west
Iran. In this study , water samples during 12 months were collected to
examine Biochemical Oxygen Demand (BOD) and Dissolved
Oxygen(DO) as a criterion for evaluation of water pollution of the
reservoir. In summary ,the study has shown Maximum, average and
minimum levels of BOD have observed 25.9 ,9.15 and 2.3 mg/L
respectively and statistical parameters of data such as standard
deviation , variance and skewness have calculated 7.88, 62 and 1.54
respectively. Finally the results were compared with Iranian national
standards. Among the analyzed samples, as the maximum value of
BOD (25.9 mg/L) was observed at the May 2010 , was within the
maximum admissible limits by the Iranian standards.
Abstract: Control chart pattern recognition is one of the most important tools to identify the process state in statistical process control. The abnormal process state could be classified by the recognition of unnatural patterns that arise from assignable causes. In this study, a wavelet based neural network approach is proposed for the recognition of control chart patterns that have various characteristics. The procedure of proposed control chart pattern recognizer comprises three stages. First, multi-resolution wavelet analysis is used to generate time-shape and time-frequency coefficients that have detail information about the patterns. Second, distance based features are extracted by a bi-directional Kohonen network to make reduced and robust information. Third, a back-propagation network classifier is trained by these features. The accuracy of the proposed method is shown by the performance evaluation with numerical results.
Abstract: Proactive coping directed at an upcoming as opposed
to an ongoing stressor, is a new focus in positive psychology. The
present study explored the proactive coping-s effect on the workplace
adaptation after transition from college to workplace. In order to
demonstrate the influence process between them, we constructed the
model of proactive coping style effecting the actual positive coping
efforts and outcomes by mediating proactive competence during one
year after the transition. Participants (n = 100) started to work right
after graduating from college completed all the four time-s surveys
--one month before (Time 0), one month after (Time 1), three months
after (Time 2), and one year after (Time 3) the transition. Time 0
survey included the measurement of proactive coping style and
competence. Time 1, 2, 3 surveys included the measurement of the
challenge cognitive appraisal, problem solving coping strategy, and
subjective workplace adaptation. The result indicated that proactive
coping style effected newcomers- actual coping efforts and outcomes
by mediating proactive coping competence. The result also showed
that proactive coping competence directly promoted Time1-s actual
positive coping efforts and outcomes, and indirectly promoted Time
2-s and Time 3-s.
Abstract: In the first part of the research work, an electrolyzer (10.16 cm dia and 24.13 cm height) to produce hydrogen and oxygen was constructed for single slice O2/H2 fuel cell using cation exchange membrane. The electrolyzer performance was tested with 23% NaOH, 30% NaOH, 30% KOH and 35% KOH electrolyte solution with current input 4 amp and 2.84 V from the rectifier. Rates of volume of hydrogen produced were 0.159 cm3/sec, 0.155 cm3/sec, 0.169 cm3/sec and 0.163 cm3/sec respectively from 23% NaOH, 30% NaOH, 30% KOH and 35% KOH solution. Rates of volume of oxygen produced were 0.212 cm3/sec, 0.201 cm3/sec, 0.227 cm3/sec and 0.219 cm3/sec respectively from 23% NaOH, 30% NaOH, 30% KOH and 35% KOH solution (1.5 L). In spite of being tested the increased concentration of electrolyte solution, the gas rate does not change significantly. Therefore, inexpensive 23% NaOH electrolyte solution was chosen to use as the electrolyte in the electrolyzer. In the second part of the research work, graphite serpentine flow plates, fiberglass end plates, stainless steel screen electrodes, silicone rubbers were made to assemble the single slice O2/H2 polymer electrolyte membrane fuel cell (PEMFC).
Abstract: Short term electricity demand forecasts are required
by power utilities for efficient operation of the power grid. In a
competitive market environment, suppliers and large consumers also
require short term forecasts in order to estimate their energy
requirements in advance. Electricity demand is influenced (among
other things) by the day of the week, the time of year and special
periods and/or days such as Ramadhan, all of which must be
identified prior to modelling. This identification, known as day-type
identification, must be included in the modelling stage either by
segmenting the data and modelling each day-type separately or by
including the day-type as an input. Day-type identification is the
main focus of this paper. A Kohonen map is employed to identify the
separate day-types in Algerian data.
Abstract: Integration of system process information obtained
through an image processing system with an evolving knowledge
database to improve the accuracy and predictability of wear debris
analysis is the main focus of the paper. The objective is to automate
intelligently the analysis process of wear particle using classification
via self-organizing maps. This is achieved using relationship
measurements among corresponding attributes of various
measurements for wear debris. Finally, visualization technique is
proposed that helps the viewer in understanding and utilizing these
relationships that enable accurate diagnostics.