Cagla Nur Oksuz

Software Engineer with a Bachelor's degree in Computer Science and expertise in software development, quality control, and machine learning. Proficient in analyzing complex problems and designing algorithms using Python, TensorFlow, OpenCV, and SQL. Experienced in collaborating with cross-functional teams throughout the project lifecycle. Currently employed at Flex Inc., demonstrating a strong commitment to delivering efficient solutions.

The project’s focus was to develop a new object detection and counting techniques and model in retail business. Used the test data and the model's predictions, mathematical equations and the necessary graphics were created and interpreted by applying Python, TensorFlow API, Fast R-CNN, RPN, ROI Pooling, Label Map, Frozen Model, Eval., 

• Algorithms applied: Fast R-CNN was used to make changes to th boxing function in TensorFlow API to control the object counting process. CSV format data file used for object counting. Actual object counts were thrown into the array. The object numbers predicted by the test results model were thrown into other arrays. Then, with the help of these sequences, the model accuracy was mathematically calculated.
• Machine Learning: A training of 2000 epochs has been carried out. Frozen model was used for the test and several proofs were made. It was concluded that the object detection and counting model worked and proofed the equation for all 4 objects with overall success rate of 83%.

High accuracy of this model offered to be a new faster solution in many industries including retail businesses.

Project was continuation of object detection with TensorFlow and the main objective of the project was to distinguish specific objects by applying TensorFlow, TFLearn, PIL, Numpy, OpenCv, Matplotlib.

Anti-Aircraft Game (console app).

 Disease survey app.

Created fuzzy logic model with blood dataset, including MAPE calculation to analyze blood donation ability.