An Approach for Regression Analysis with Intuitionistic Fuzzy Random Data
Project no. 22661 · Status: Kapanmış · 2021
Summary
<p>Nowadays, computer technologies have become an indispensable part of human life. Even simple household appliances used in daily life contain various computer-aided systems. In addition, many complex systems and structures still perform in a computer-aided manner. Such prevalence is mainly based on scientific developments in Artificial Intelligence and machine learning. Fuzzy Logic, which is one of the backbone issues of Artificial Intelligence, is a scientific area put forward by L.A. Zadeh. With the concept of Fuzzy Logic, it has been possible for computer-aided systems to evaluate like humans. With the contribution of K. Atanassov in 1986, this development continued with the concept of Intuitionistic Fuzzy Logic.</p><p><br></p><p>Intuitionistic Fuzzy Logic provides solutions especially when datasets contain incomplete information. Data sets are important when working with Machine Learning techniques. At the stage of applying Regression Analysis, which is a Machine Learning technique, the data at hand should provide some assumptions and should not contain uncertainty. Fuzzy or Intuitionistic Fuzzy techniques can be used when these conditions are not met.</p><p><br></p><p>In this project, a Regression Model will be developed that can be applied if such data are available. This model will be expressed both theoretically and its performance will be measured through various data sets. In the model development phase, a new distance formula, which will basically be defined between Intuitionistic Fuzzy Numbers, will be effective and decisive. Thanks to this distance formula, a flexible, adjustable Regression Model that can reflect the user's views on the process will be created.</p>