Tuncay Yigit

Web-based learning object selection software using analytical hierarchy process

The concepts of sustainability and reusability have great importance in engineering education. In this context, metadata provides reusability and the effective use of Learning Objects (LOs). In addition, searching the huge LO Repository with metadata requires too much time. If the selection criteria do not exactly match the metadata values, it is not possible to find the most appropriate LO. When this situation arises, the multi-criteria decision making (MCDM) method can meet the requirements. In this study, the SDUNESA software was developed and this software allows for the selection of a suitable LO from the repository by using an analytical hierarchy process MCDM method. This web-based SDUNESA software is also used to store, share and select a suitable LO in the repository. To meet these features, the SDUNESA software contains Web 2.0 technologies such as AJAX, XML and SOA Web Services. The SDUNESA software was especially developed for computer engineering education. Instructors can use this software to select LOs with defined criteria. The parameters of the web-based SDUNESA learning object selection software that use the AHP method are defined under the computer education priorities. The obtained results show that the AHP method selects the most reliable learning object that meets the criteria.

Constraint-Based School Timetabling Using Hybrid Genetic Algorithms

In this paper, a hybrid genetic algorithm (HGA) has been developed to solve the constraint-based school timetabling problem (CB-STTP). HGA has a new operator called repair operator, in addition to standard crossover and mutation operators. A timetabling tool has been developed for HGA to solve CB-STTP. The timetabling tool has been tested extensively using real-word data obtained the Technical and Vocational High Schools in Turkey. Experimental results have presented that performance of HGA is better than performance of standard GA.

An educational tool for controlling of SRM

This article introduces an educational tool for a switched reluctance motor (SRM) drive system. It is prepared for undergraduate and graduate level students. Classical PI and Genetic PI controllers are used in SRM drive system. The Genetic PI controller was applied to the speed loop, replacing the classical PI controller. The tool software was implemented using C++ Builder on a PC. It has flexible structure and graphical interface. The students can be easily establishing a thorough understanding of both classical PI and genetic PI controller for a SRM drive system. The education tool allowed the student to interact with the SRM drive system and it is using controllers. Then it is responses on a dynamic and instantaneous basis under different operating conditions.

Mobile indoor navigation system in iOS platform using augmented reality

Usage areas of mobile phones have increased in the last 10 years. Although there have been improvements in many areas, most of the developments are in the field of positioning systems. Although the peoples lives continue in indoor environments, location-based information system receives data from the satellites, which can detect a persons location in outdoor areas. In indoor areas, satellite signals can cause false information or can be interrupted by the objects in the area. In this study, an indoor navigation system has been designed and developed that only uses the accelerometer, the camera and the compass components on the phone and does not require satellite signals for positioning. To provide independence from the map in this application, augmented reality is applied during the routing process by utilizing built-in camera of the phone and no map is used.

Implementation of wavelet transform extrasystole heart sound with convolution method for feature extraction

In this study, discrete wavelet transform (DWT) has been implemented using convolution method with purpose of feature extraction of extra systole heart disorder in heart sound database towards classification. Dual down sampling has been applied for DWT method and reflectance has been applied to avoid boundary effect. Approximation and detail coefficients that obtained using DWT implementation has been compared with coefficients that obtained using MATLAB program using regression analysis. As result of the regression, the mean error squared (RMSE) has been found ideally as 0.00002515 and the R-squared value of 1.

Classification of extrasystole heart sounds with MFCC features by using Artificial Neural Network

In this study, classification of Normal and Extra systolic heart sounds (HS) have been carried out using in PASCAL Heart Sounds (HS) data base. The extrasystole is the HS that is produced by performing an extra beat in each heart cycle, unlike the heartbeat normal cycle. It can be felt by people as palpitations. Occurrence of these sounds in certain age groups may be the indication of tachycardia. In this study, firstly HS have been normalized at first. Then an elliptic filter has used for noise reduction. HS features have been obtained using Mel-Frequency Cepstrum Coefficients. These features have classified using Artificial Neural Network. In this study, 45 extra systoles heart sounds have used. 30 of them have been used as training data for classification while remaining 15 ones have been used for the test. Certainty, sensitivity, accuracy values have been calculated using confusion matrix. Classification success has been calculated as 90%.

A hybrid AHP-GA method for metadata-based learning object evaluation

A wide variety of demand in e-learning and web-based learning caused a new approach in e-content presentation. In order to accomplish these demands, learning object repositories (LORs) were developed. LORs have many learning objects (LOs) that are used to produce different types of e-content. When there are many LOs in LORs, the evaluation and selection of them become a subjective and time-consuming process. Thus, selecting the most suitable and best qualified LO is considered as a multi-criteria decision-making (MCDM) problem. In this study, a hybrid analytic hierarchy process genetic algorithm (AHP-GA) method was developed for the evaluation of LOs from web-based Intelligent Learning Object Framework (Zonesa) LOR. This proposed hybrid system was used in a real case study and the results demonstrated that the proposed system can be used effectively by both users and machines to produce content by the help of LO metadata.

Using the metaheuristic methods for real-time optimisation of dynamic school bus routing problem and an application

The vehicle routing problem (VRP) is an optimisation issue that has been studied for more than 50 years with its numerous subfields. The optimisation of VRP over distribution and transportation systems leads to significant gains in cost and time. There are many metaheuristic methods developed for the solution of the problem; and it was observed that metaheuristic methods prove to produce more successful results compared to common heuristic methods. In this study, a mobile-supported visual application was developed using ant colony optimisation (ACO) and genetic algorithm (GA), which are among the metaheuristic methods for the dynamic school bus routing problem (DSBRP), one of the sub-problems of VRP. The ACO and GA methods were utilised via the application for bus routes of a school located in the province of Ankara and the performance of these methods were compared through the obtained results. It was observed that time and distance values of the routes of current school bus routes may be improved by these two methods.

Resource allocation using ANN in LTE

LTE is the 4th generation wireless network technology, which provides flexible bandwidth, higher data speeds and lower delay. Difficulties may be experienced upon an increase in the number of users in LTE. The objective of this study is to ensure a faster solution to any such resource allocation problems which might arise upon an increase in the number of users. A fast and effective solution has been obtained by making use of Artificial Neural Network. As a result, fast working artificial intelligence methods may be used in resource allocation problems during operation.

Expert system software for domestic animals

Today, information technology has been used frequently in the veterinary field. Storage, information gathering and interpretation processes are realized by transferring the related field data to digital medium. In this context, expert systems provide important implications for symptom-disease relationship. In the study, the symptoms obtained from the pets are evaluated by the expert system and the inference is made about the diagnosis of the disease. Expert system software developed with ASP.NET MVC has responsive mobile and web interface. In the software, the user obtains information about the possible diseases by entering symptoms from the interface. With a web based software developed in a dynamic structure, it is possible to update symptom and illness data of specialist veterinarian about pets. Thus, new diseases and symptoms can be transferred to the written form. Software and veterinary expert veterinary information may be used by other veterinarians in the area concerned. The software is being used to cover other animals except cats / dogs.