CV

PDF Version

Education

• Doctor of Philosophy - Electrical and Computer Engineering (4.00/4.00), 2024-2029
  Purdue University - The Elmore Family School of Electrical and Computer Engineering - Indiana, US.
  Advisor: Prof. Abolfazl Hashemi

• Master Of Science - Electrical and Electronics Engineering (3.91/4.00), 2021-2024
  Bilkent University - Faculty of Engineering - Ankara, Turkey.
  Advisor: Prof. Suleyman Serdar Kozat

• Bachelor of Science - Electrical and Electronics Engineering (3.61/4.00), 2017-2021
  Middle East Technical University - Faculty of Engineering - Ankara, Turkey.

Publications

Journal Papers

1. S. F. Tekin, A. Fazla, S. S. Kozat, “Numerical Weather Forecasting using Convolutional-LSTM with Attention and Context Matcher Mechanisms”, IEEE Transactions on Geoscience and Remote Sensing, 2024.
   Available: https://arxiv.org/abs/2102.00696
   Code: https://github.com/sftekin/spatio-temporal-weather-forecasting

2. A. Fazla, M. E. Aydin, S. S. Kozat, “Time-Aware and Context-Sensitive Ensemble Learning for Sequential Data”, IEEE Transactions on Artificial Intelligence, 2023.
   Available: https://doi.org/10.1109/TAI.2023.3319308
   Code: https://github.com/ardafazla/context-time-aware-ensemble

3. M. E. Aydin, A. Fazla, S. S. Kozat, “Hybrid State Space-based Learning for Sequential Data Prediction with Joint Optimization”, arXiv preprint arXiv:2309.10553, 2023.
   Available: https://arxiv.org/abs/2309.10553
   Code: https://github.com/mustafaaydn/lstm-sx

4. A. Fazla, M. E. Aydin, S. S. Kozat, “Joint Optimization of Linear and Nonlinear Models for Sequential Regression”, Digital Signal Processing, Elsevier, 2022.
   Available: https://doi.org/10.1016/j.dsp.2022.103802
   Code: https://github.com/ardafazla/jointoptimization

Conference Papers

1. K. G. Ince, A. Koksal, A. Fazla, A. A. Alatan, “Semi-Automatic Annotation for Object Detection”, Proceedings of the IEEE/CVF International Conference on Computer Vision, 1233-1239, 2021.
   Available: https://doi.org/10.1109/ICCVW54120.2021.00143

Academic Awards

The Elmore Family School of Electrical and Computer Engineering, Purdue University

2024-2029

• Awarded the highly competitive and prestigious Elmore Fellowship Award for my graduate studies, given only to the most outstanding applicants.

Turk Telekom & Information and Communication Technologies Authority

2023-2024

• Awarded 5G and Beyond Joint Graduate Support Programme, a merit-based fellowship of monthly stipend during M.Sc.

Scientific and Technological Research Council of Turkey

2021-2023

• Awarded Directorate of Science Fellowships and Grant Programme, a merit-based monthly stipend and accommodation support during M.Sc.

• Received the 373rd rank among 2M high school graduates in University Entrance Examination.

• Received the 11th rank among 0.2M university graduates in ALES (National GRE).

Academic Duties

Reviewer Duties

Conferences: ICLR25, AAAI25
Journals: IEEE TAI25, IEEE IOT24

Teaching Assistantship

Electrical and Electronics Engineering, Bilkent University, 2021-2024

• EEE321 Signals and Systems
• EEE202 Circuit Theory
• MATH255 Probability and Statistics

Relevant Coursework

• Optimization for Deep Learning (4.0/4.0)
• Statistical Learning and Data Analytics (4.0/4.0)
• Random Processes (4.0/4.0)
• Deep Generative Networks (4.0/4.0)
• Data Structures (4.0/4.0)
• Digital Signal Processing (4.0/4.0)

Research Projects

Hourly Wind Energy Prediction

Bilkent University, 2024

• Predicted the hourly energy data of multiple wind turbines in Turkey and Europe, showing chaotic and nonstationary behavior in multiple regions.
• Developed a large-scale ML framework consisting of various adaptive feature construction and selection methods with models such as deep-learning (NN, ConvLSTM), machine learning (LightGBM, RandomForest), statistical (SARIMAX) models and state-of-the-art ensembling techniques. Increased the short-term forecasting accuracy over state-of-the-art models by an average of 21\% and long-term forecasting accuracy by an average of 8\%.

Natural Gas Demand Prediction

Bilkent University, 2023

• Predicted the daily data of international natural gas demand in Turkey, which contains nonstationary multivariate time series from multiple sources.
• Developed a large-scale ML framework specifically tailored towards a certain customer profile, where the predictions are obtained for the given data based on key indicators that are automatically extracted from the data itself and/or the user-given side information data. The constructed framework was deployed to various natural gas production companies.

Small-Target Detection & Tracking

Middle East Technical University, 2021

• Developed an application in Python environment for real-time small target tracking by leveraging temporal data derived from the output of object detectors, with a superior performance over the widely-known YOLOv3 algorithm by 6\%.
• Constructed a novel multi-hypothesis tracking algorithm with an annotation tool for the semi-automatic correction of mislabeled UAV data. The deployed product decreased the total mislabeled data in various datasets by an average of 13\%.

Additional Projects

Multi-modal Photo Upsampling via Latent Space Exploration of StyleGAN

Bilkent University, Deep Generative Netwoks Course Project, 2023

• Adapted the PULSE algorithm for multi-modal image super-resolution by incorporating a variational autoencoder to enhance network diversity.

Image Captioning based on LSTM with Transfer Learning

Neural Networks Course Project, 2022

• Constructed a framework, which generates meaningful sentences describing a given image (Image Captioning).
• Employed state-of-the-art models such as ResNet, VGG-16 and InceptionV3, where LSTM is used as the main learning model for transfer learning.
• Used BERT transformers for word embedding.

Software Skills

• Pyhton: Professional research and industrial experience based on machine learning. High knowledge and experience on libraries such as PyTorch, Tensorflow, Pandas, Numpy and Scikit-Learn. I have experience with Docker and Git, via industrial projects.
• MATLAB: Professional research experience in signal processing and computer vision. Used for projects during my undergraduate courses, undergraduate research, and internships.
• R: Intermediate level experience, employed during research based on time-series forecasting using statistical models.
• C/C++: Intermediate experience in Arduino and various undergraduate projects.

Languages

Turkish: Native
English: Conversationally Fluent [TOEFL IBT 108/120 (October 22, 2024)]