Publications

4th International Artificial Intelligence and Data Science Congress ICADA 2024

The Prediction of Technology Companies Mergers and Acquisitions Through Machine Learning Methods

In this study, merger and acquisition strategies of leading technology companies in 2021 were predicted and analyzed using machine learning methods. The dataset used in the study includes acquisition transactions carried out by technology companies such as Microsoft, Google, IBM, HP, Apple, Amazon, Facebook, Twitter, eBay, Adobe, Citrix, Redhat, BlackBerry, and Disney, which are pioneers in the industry. The data was compiled from reliable sources such as Wikipedia, TechCrunch, and CrunchBase. Accordingly, various attributes such as the date, year, month of each acquisition transaction, the name of the acquired company, the acquisition cost or value, the business use case of the acquisition, and the country of acquisition were included. Artificial Neural Networks, k-Nearest Neighbors, Support Vector Machines, Logistic Regression, Random Forest, Naive Bayes, and AdaBoost machine learning algorithms were employed based on these attributes. The findings of the study present a general overview of merger and acquisition activities in the technology sector in 2021 by revealing trends and significant patterns in the analyzed dataset. The study is expected to serve as a valuable resource for researchers and industry professionals who aim to understand the strategic decisions of technology companies, evaluate competition in the sector, and anticipate future developments.

---

3rd International Conference on Engineering, Natural and Social Sciences ICENSOS 2024

Diabetes Risk Prediction Using Machine Learning Methods

The aim of this study is to predict diabetes risk at an early stage using machine learning methods. In the research, models were developed to categorize diabetes risk by analyzing demographic and clinical features. These models will assist healthcare providers in identifying potential diabetes cases in advance and intervening accordingly, which will help reduce treatment costs and improve patient care quality. Individuals’ diabetes risk profiles were created based on symptoms such as polyuria, polydipsia, and sudden weight loss, increasing the possibility of early intervention. The study developed models using various machine learning algorithms such as Logistic Regression, Decision Trees, Support Vector Machines (SVM), and Random Forest. The performances of these models were evaluated and compared. The results revealed that the model with the highest success was the Support Vector Machines (SVM) model.