AdaCL
![](/our-papers/adacl/teaser.png)
Hi, I am a PhD candidate under the supervision of Joaquin Vanschoren at the department of Mathematics and Computer Science in Eindhoven University of Technology, the Netherlands.
Continual Learning, Computer Vision, Deep Learning.
Hi again, I am working on Continual Learning. Continual Learning is a specific field of AI where we teach our models with sequential data. What I mean by 'sequential data' is this: instead of assuming we have access to lots of data at once, we assume that the data comes to us part by part. We generally name this streaming data as "task". Our goal is to keep training our model with upcoming data (yes basically finetuning! ðŸ¤). But there is a big problem here! As humans we also have this problem when learning new things. Can you guess what that is? It is of course the forgetting issue 🙃. When we keep learning the new task we unfortunately forget previous things. So, in continual learning we try to solve this forgetting issue while learning to new things. By saying "we" I mean many enthusiastics in this field. If you find this problem interesting, like I do, and would like to chat and even collaborate on this, I am looking forward to meet with you 🤓. You can reach me via the social links. See ya!
07/2024, I have attended a summer school on Generative Modelling and here is my high level notes: GeMMS.
03/2024, We have a new hot paper called "FOCIL: Finetune-and-Freeze for Online Class Incremental Learning by Training Randomly Pruned Sparse Experts": FOCIL
11/2023, Our paper "Continual Learning with Dynamic Sparse Training: Exploring Algorithms for Effective Model Updates" has been accepted for oral presentation at the CPAL.
10/2023, We have presented AdaCL in virtual: AdaCL:Adaptive Continual Learning
09/2023, Our paper "AdaCL: Adaptive Continual Learning" has been accepted for oral presentation at the CLAI.
04/2023, Our model "Condition Monitoring and Predictive Maintenance in Railways" is granted to international patent.
10/2022, Our paper "Automated Machine Learning Approach in Material Discovery of Hole Selective Layers for Perovskite Solar Cells" has been published in the Energy Technology.
11/2021, Our paper "Predicting Perovskite Bandgap and Solar Cell Performance with Machine Learning" has been published in the Solar RRL.
05/2021, Our paper "A Machine Learning Approach for Metal Oxide Based Polymer Composites as Charge Selective Layers in Perovskite Solar Cells" has been published in the ChemPlusChem.
08/2020, Our paper "Comparison of Machine Learning Models on Performance of Single- and Dual-Type Electrochromic Devices" has been published in the ACS Omega.
Yildirim, M. O., Gok Yildirim, E. C., Sokar, G., Mocanu, D. C., Vanschoren, J. Continual Learning with Dynamic Sparse Training: Exploring Algorithms for Effective Model Updates. In CPAL, 2024. paper
Yildirim, M. O., Gok Yildirim, E. C., Eren, E., Huang, P., Haris, M. P., Kazim, S., Vanschoren, J., Oksuz, A.U., Ahmad, S. Automated Machine Learning Approach in Material Discovery of Hole Selective Layers for Perovskite Solar Cells. In Energy Technology, 2023. paper
Gok, E. C., Yildirim, M. O., Haris, M. P., Eren, E., Pegu, M., Hemasiri, N. H., Huang, P., Kazim, S., Oksuz, A.U., Ahmad, S. Predicting perovskite bandgap and solar cell performance with machine learning. In Solar RRL, 2022. paper
Yildirim, M. O., Gok, E. C., Hemasiri, N. H., Eren, E., Kazim, S., Oksuz, A. U., Ahmad, S. A Machine Learning Approach for Metal Oxide Based Polymer Composites as Charge Selective Layers in Perovskite Solar Cells. In ChemPlusChem, 2021. paper
Gök, E.C., Olgun, M.O. SMOTE-NC and gradient boosting imputation based random forest classifier for predicting severity level of COVID-19 patients with blood samples. In Neural Computing & Applications, 2021. paper
Gok, E. C., Yildirim, M. O., Eren, E., Oksuz, A. U. Comparison of Machine Learning Models on Performance of Single-and Dual-Type Electrochromic Devices. In ACS Omega, 2020. paper