Makale Koleksiyonu

Bu koleksiyon için kalıcı URI

https://hdl.handle.net/20.500.12452/6253

Listeleniyor 1 - 7 / 7

Development of a Prototype Using the Internet of Things for Kinetic Gait Analysis
(2018) Çalişkan, Muhammet; Tümer, Abdullah Erdal; Şengül, Sümeyra Büşra
The proliferation of mobile devices and the gradual development of technology have led to the emergence of the concept ofInternet of Things. The IoT has led to an increase in the work done especially on the medical field. The beginning of the reasons forusing the IoT in medical studies is to be able to detect and display instantaneous changes that physicians cannot even observe. The aim ofthis study is to contribute to the recovery of The Gait Analysis from the constraints such as cost, expert necessity and difficulty ofmeasuring the natural walking, and to make Gait Analysis widespread by realizing the more cost-effective. Another contribution of thestudy is to determine the high pressure points in the foot base and to prevent the loss of tissue in the feet by being produced theappropriate base for the patient. In the study, a prototype placed inside the shoe with the internet of things is developed to monitor thepressure distribution of the foot base. The prototype consists of a thin, flexible insole that collects analog data from the 32 sensors andtransmits it wirelessly to mobile or PC via Bluetooth technology. The developed software of the prototype shows the pressure in everysensor on the floor and draws the walking chart. The accuracy and reliability of the prototype are assessed by pre-experimentalmeasurements. The prototype is tested on 14 male and 4 female participants. The prototype is tested on people at 105 kg and below.
Human Gender Prediction on Facial Images Taken by Mobile Phone using Convolutional Neural Networks
(2018) Hacıbeyoğlu, Mehmet; İbrahim İbrahim, Mohammed Hussein
The interest in automatic gender classification has increased rapidly, especially with the growth of online social networkingplatforms, social media applications, and commercial applications. Most of the images shared on these platforms are taken by mobile phonewith different expressions, different angles and low resolution. In recent years, convolutional neural networks have become the mostpowerful method for image classification. Many researchers have shown that convolutional neural networks can achieve better performanceby modifying different network layers of network architecture. Moreover, the selection of the appropriate activation function of neurons,optimizer and the loss function directly affects the performance of the convolutional neural networks. In this study, we propose a genderclassification system from facial images taken by mobile phone using convolutional neural networks. The proposed convolutional neuralnetworks have a simple network architecture with appropriate parameters can be used when rapid training is needed with the amount oflimited training data. In the experimental study, the Adience benchmark dataset was used with 17492 different images with different genderand ages. The classification process was carried out by 10-fold cross validation. According the experimental results, the proposedconvolutional neural networks predicted the gender of the images 98.87% correctly for training and 89.13% for testing.
Ses yalıtımında ses azaltım indisi modellerinin karşılaştırmalı olarak incelenmesi
(2016) Aksoylu, Ceyhun; Mendi, Şekip Engin; Arda, Söylev
Teorik hesaplamalara dayalı olarak geliştirilmiş ses yalıtım modellerinin amacı, deneysel olarak malzemeye ait "Ses Azaltım İndisi (R)" değerlerinin bulunması sırasında harcanan zaman, emek ve maliyeti azaltmaktır. Bu çalışmada, literatürde var olan simülasyon modelleri ile ticari ses yalıtım programları kullanılarak, yapı malzemelerinin R değerleri araştırmacılar tarafından daha önce yapılan deneysel verilerle 11 farklı malzeme için karşılaştırmalı olarak incelenmiştir. Deneyler Bastian, Akuzoft, Insul ve dBKAisla modelleri kullanılarak yapılmıştır. Modeller ISO 12354, ISO 10140 ve ISO 717 standartlarında 1/3 oktav bant analiz yapmaktadır. Kullanılan bu modellerde literatürdeki temel hesaplama formülleri, tek tabakalı paneller için R hesabında kullanılması gereken hesap yöntemleri, kullanılan panel boyutları ile malzemeye ait kalınlık, yoğunluk, elastisite modülü, porozite, iç kayıp faktörü ve sesin havadaki hızı dikkate alınmıştır. Bu sayede, farklı frekanslara karşılık gelen R değerleri deneysel çalışmalarla bulunmuş malzemelerin, farklı ses yalıtım modelleri kullanılarak simülasyonları yapılmış, sonuçları karşılaştırılarak kullanılan ses yalıtım modellerinin etkinlikleri belirlenmiştir. Analizler sonucunda, kullanılan modellerin farklı malzemeler için R ve R'ye bağlı olarak malzemelerin performansını gösteren tek dereceli "Ağırlıklı ses azaltım indisi (Rw)" bakımından etkinlikleri saptanmış, doğruluk değerleri hesaplanmıştır.
Biyomedikal veri kümeleri ile makine öğrenmesi sınıflandırma algoritmalarının istatistiksel olarak karşılaştırılması
(2014) Karakoyun, Murat; Hacibeyoğlu, Mehmet
Günümüzde bilişim teknolojileri hemen hemen her alanda kullanılmaktadır. En çok kullanılan alanlardan bir tanesi de sağlık sektörüdür. Dijital hastane sistemlerinin kullanılmasıyla birlikte hasta verileri artık bilgisayar ortamında saklanmakta ve böylelikle biyomedikal veri kümeleri oluşmaktadır. Boyut olarak çok büyük olan bu veri kümelerinin bir insan tarafından analiz edilmesi ve yorumlanması çok zordur. Bunun için bilgisayar mühendisliği çalışma alanlarından biri olan makine öğrenmesi algoritmaları kullanılır. Bu çalışmada 6 tane makine öğrenmesi algoritmalarının başarımları 9 farklı biyomedikal veri kümesi üzerinde test edilmiştir ve elde edilen sonuçlar istatistiksel olarak karşılaştırılmıştır. Deneysel ve istatistiksel sonuçlar birlikte incelediğinde küçük ve orta büyüklükteki biyomedikal veri kümeleri için Yapay Sinir Ağları algoritması sınıflandırma başarımı açısından ve K- en Yakın Komşu algoritması ise çalışma zamanı açısından daha başarılı olmuştur. Bu çalışmanın bir bölümü ASYU 2014/İzmir sempozyumunda bildiri olarak sunulmuştur.
Improved Nelder-Mead Optimization Method in Learning Phase of Artificial Neural Networks
(2018) Koçer, Hasan Erdinç; Merdan, Mustafa; Ibrahim, Mohammed Hussein
Artificial neural networks method is the most important/preferred classification algorithm in machine learning area. The weightson the nets in artificial neural directly affect the classification accuracy of artificial neural networks. Therefore, finding optimum values ofthese weights is a difficult optimization problem. In this study, the Nelder-Mead optimization method has been improved and used fortraining of artificial neural networks. The optimum weights of artificial neural networks are determined in the training stage. Theperformance of the proposed improved Nelder-Mead-Artificial neural networks classification algorithm has been tested on the mostcommon datasets from the UCI machine learning repository. The classification results obtained from the proposed improved Nelder-Mead-Artificial neural networks classification algorithm are compared with the results of the standard Nelder-Mead-Artificial neural networksclassification algorithm. As a result of this comparison, the proposed improved Nelder-Mead-Artificial neural networks classificationalgorithm has given best results in all datasets.
An Artificial Neural Network Model for Wastewater Treatment Plant of Konya
(2015) Tümer, Abdullah Erdal; Edebali, Serpil
In this study, modelling of Konya wastewater treatment plant was studied by using artificial neural network with different architectures in Matlab software. All data were obtained from wastewater treatment plant of Konya during daily records over four month. Treatment efficiency of the plant was determined by taking into account of input values of pH, temperature, COD, TSS and BOD with output values TSS. Performance of the model was compared via the parameters of Mean Squared Error (MSE), and correlation coefficient (R). The suitable architecture of the neural network model is determined after several trial and error steps. According to the modelling study, the ANN can predict the plant performance with correlation coefficient (R) between the observed and predicted output variable reached up to 0.96.
Artificial Neural Network Models for Predicting the Energy Consumption of the Process of Crystallization Syrup in Konya Sugar Factory
(2017) Tümer, Abdullah Erdal; Koç, Bilgen Ayan; Koçer, Sabri
In this study, a model has been developed from the sugar production process stages in Konya Sugar Factory using artificial neural networks to estimate the energy consumption of the process of crystallization syrup. Model developing specific enthalpy, mass and pressure as input layer parameters and consumption energy as output layer was used. 124 different data are taken from Konya Sugar Factory during January 2016. Feed-forward backpropagation algorithm was used in the training phase of the network. Learning function LEARNGDM and the number of hidden layer kept constant as 2 and transfer functions are modified. To find the most optimal model, 27 artificial neural networks with different architectures have been tested. 2-5-1 network architecture was determined as the best suitable network architecture and transfer function is determined logsig function as the optimal transfer function. Optimum results of the model taken in the coefficient of determination was found R 0.98 neural network training, testing and validate was also found to be R 0.98, the performance of the network for not shown data to network was found R0.99

Güncel Gönderiler