| Artificial Intelligence Research |
 |
|
What is Artificial Intelligence?
|
|
According to Wikipedia, Artificial intelligence (AI) is intelligence demonstrated by machines, unlike the natural intelligence
displayed by humans and animals, which involves consciousness and emotionality. The distinction between the former and the
latter categories is often revealed by the acronym chosen. 'Strong' AI is usually labelled as Artificial General
Intelligence while attempts to emulate 'natural' intelligence have been called Artificial Biological Intelligence.
|
|
Our Funded Research Projects
|
- Digital Soil Texture Analyzer, TUBITAK, 2018-2020.<click for details>
In this project, a new system based on tracking of settling particles in water with ultrasound sensors is developed and the findings are presented in detail. This new system, called the Ultrasound Soil Texture Analyzer (USTA) based on ultrasound penetration consists of a 3D-printed container, a pair of ultrasound sensors working at 1 MHz frequency facing each other from the outer surface of the container, an oscilloscope and a computer. Different experiments are carried out on the design of the container, the frequencies and positions of the sensors used, temperature change, the way of reciprocating, and the effects of sodium-hexametaphosphate solution on soils, in order to explore the capabilities and responses of this new device.
- Live Turkish Dictionary, TUBITAK, 2016-2019 <click for details>
The project is based on a system design that analyzes texts collected from various internet news sites and
performs the following operations: Identifying the new words entering Turkish, Determining a numerical relationship from the
union of these new words with known words, Using this new words and detected relationships to improve the synset space
calculated from a Turkish WordNet. As a result of the project; 3 journal articles were prepared, 4 symposium procedings
were presented, 2 students were funded throughout their education duration.
- Learning Word-Vector Quantization: A Study In Morphological Disambiguation of Turkish, CU-BAP, 2015-2020
It introduces a new classifier named Learning Word-vector Quantization (LWQ) to solve morphological
ambiguities in Turkish, which is an agglutinative language. Also, a new and morphologically annotated
corpus, and then its datasets are prepared with a series of processes.
- Estimation of traditional ECG with signals measured from the wrist, CU-BAP 2015-2017.
In this study, only left-arm ECG signals were collected and it is aimed to determine cardiac rhythm with high accuracy as an example application. Dual-arm measurements are also taken on the subjects, simultaneously with single-arm signals, to obtain the signal known as Lead I in conventional ECG measurement and to use it as a reference. In these collected signals, we focus on primarily eliminating noise and then selecting the most significant frequency sub-bands. Finally, the heart rate is determined by an algorithm that counts heartbeats, and the heart rate of the double-arm beats collected from the same subjects is used for comparison.
- Prediction of Soil Mixture by Using Intelligent Algorithms, CU-BAP, 2014-2016.
In this study, it is aimed to make soil texture analysis with an electronically supported system. The signals obtained in the experiments based on the monitoring of a container full of water-soil solution with the embedded system and light sensors connected to it are recorded on a computer and these signals are analyzed by machine learning methods.
- A Model Transforming A Problem Based on Chance Problem into Perfect Information and Its Computer Simulation, TUBITAK, 2007.
In this project, a simulation of no-chance backgammon, a new competent knowledge game, based on minimax and fuzzy logic, is made. First, the some statistics of the game tree is found by taking into account the search depth, the number of stones and gates in the game. After that, an evaluation function is created intuitively and observationally. As a result, a game application has been prepared using a fuzzy inference system with 17 rules and the minimax technique. Click for no-chance backgammon application for Windows.
|
|
Graduate Prospective Students
|
|
Candidate students must have a bachelor's degree in computer science. Apart from that, preferably, it is expected to have experience in Python coding and to have taken Natural Language Processing and Machine Learning courses in undergraduate education.
|
|
Courses
|
|
|
|
Thesis
|
|
You can find my students' thesis on this page by filtering supervisor.
|
|
Publications
|
My all publications are about AI. Click for Scholar.
|
|