I am someone who is always open to learning and development. I enjoy exploring new things, such as traveling, theater, piano, and cooking. I am also curious and interested in emerging technologies, particularly in the field of artificial intelligence. I recognize the importance of data in our lives and am constantly improving my skills in AI, machine learning, computer vision, and image processing to gain a better understanding of data.
• Master's Degree
• Bachelor's Degree
• GPA: 3.01
• Associate's Degree
• GPA: 3.79 • High Honor Student
• Associate's Degree
• GPA: 3.14 • Honor Student
•Utilized advanced scripting (Python/Bash) to automate deployment and configuration of Asterisk, Suricata, Ansible, and Nginx on Debian/Ubuntu systems.
•Performed system integration across TCP/UDP/IP networks using Cisco routers/switches and Lenovo/HP servers.
•Designed and maintained secure and scalable infrastructure through Linux-based automation.
• Utilized Linux commands for system integration in the main framework.
• Technologies: Debian, Ubuntu, Pardus, NTP, Static IP, DHCP.
• Taught deep learning concepts to middle school students.
• Topics: deep learning, classification, regression, genetic algorithms, optimization.
• Developed 3D image/video processing algorithms using MediaPipe, OpenCV, Python, and PyTorch.
• Contributed to deep learning model implementations, enhancing product features.
• Tested and trained deep learning models in image segmentation, object detection, and video processing.
• Technologies: OpenCV, Python, PyTorch, CUDA, Linux.
• Designed machine learning models for object detection, image segmentation, and optical character recognition.
• Technologies: OpenCV, Python, YOLOv4, YOLOv8, U-NET.
• Brand advertisement on campus, coffee talks with the executives of the best companies.
• Received essential banking pieces training online from DenizBank officials, gaining insights into banking processes.
• My graduation project, supported by TUBITAK, aims to alleviate communication inequalities.
• It equipped me with the following skills: pose recognition, keypoint detection, real-time video prediction.
• Technologies employed include LSTM, CNN, VGG16, ResNet50, Xception, MobileNetv2, Image Processing, MediaPipe, Tensorflow, Python, and OpenCV.
In this project, I classified diabetic retinopathy images using deep learning algorithms with the APTOS dataset. I employed transfer learning to train the model, utilizing the EfficientNet B5 architecture. Prior to training, I processed the dataset to enhance image quality by reducing noise and improving clarity using techniques such as CLAHE (Contrast Limited Adaptive Histogram Equalization) and Median Blur.
In this project, I classified audio files using deep learning algorithms with the UrbanSound8K dataset. I applied Convolutional Neural Networks (CNNs) for model training. To prepare the dataset for classification and prediction, I conducted preprocessing steps, including audio signal processing with Discrete Fourier Transform (DFT) to generate spectrograms.
In my portfolio, I present the Transformers-with-FFN project, where I leveraged Vision Transformers and Feedforward Networks (FFN) to accelerate vision tasks and enhance model performance. This project features a well-structured framework, a user-friendly interface, and seamless functionalities for training, inference, and result visualization.
This project demonstrates the implementation of a Variational Autoencoder (VAE) using the Fashion-MNIST dataset for generating diverse fashion samples. The project includes an organized structure and a user-friendly interface, allowing for efficient model training, sample generation, and result visualization.
In this project, I implemented a shadow detection algorithm using mean shift and Gaussian filter techniques for enhanced accuracy. The Python script processes input images and produces a series of outputs, illustrating the various stages of shadow detection.
This project focuses on analyzing and predicting GHG emission data using machine learning regression models. We will walk through the steps of data loading, cleaning, exploratory data analysis (EDA), feature engineering, model building, and evaluation.
• This conference was organized by Google Developer Group Samsun (GDG Samsun).
• My speech at the conference focused on the current role of artificial intelligence in today's world and its impact on various industries.
• Developed clustering and classification models including SVM, XGBoost, LightGBM, and RandomForest.
• Achieved highest accuracy score with XGBoost model.
• Developed clustering and classification models including SVM, XGBoost, LightGBM, and RandomForest.
• Achieved highest accuracy score with XGBoost model.
• Developed clustering and classification models including SVM, XGBoost, LightGBM, and RandomForest.
• Achieved highest accuracy score with XGBoost model.
Object-Oriented Programming (OOP) is a fundamental paradigm that has revolutionized software development by promoting modular, reusable, and maintainable code. In this article, we delve into the core principles of OOP and explore the SOLID principles, a set of design principles that enhance the robustness and scalability of object-oriented systems. Devamını oku
Veri iletişimi, iki ya da daha fazla cihaz arasında veri paylaşımı işlemidir. Bu cihazlar, ağ aracılığıyla bağlantılı olabilirler. Veri, herhangi bir dijital ya da analog sinyal şeklinde olabilir. Veri iletişimi işlemi, doğru ve güvenilir bir şekilde gerçekleştirilmesi için bazı kurallar ve protokoller ile yönetilir. Devamını oku