Hi! I am Dovydas Joksas, aka Yoshke. I am a PhD student at University College London (UCL) where I investigate machine learning hardware. I also tutor GCSE, A-level and IBDP students in mathematics and physics.
You can download my CV from here.
PhD in Electronic Engineering (2018-2022, UCL)
Artificial neural networks (ANNs) have achieved spectacular results over the last decade, but at a cost of low efficiency. Our conventional computer hardware is not well suited to the computations that ANNs perform and so a lot of time and energy is wasted. One of the suggested alternatives is an emerging technology—memristor-based hardware. Although vastly more efficient, memristor-based implementations exhibit a greater degree of volatility and uncertainty than the digital computers that we use every day.
Using simulations, I investigate to what extent the accuracy of ANNs is affected when memristor-based hardware is used to implement them. I am also interested in whether the accuracy can be increased not by optimising device performance, but by using different software approaches. I explain the rationale behind the research in less technical terms here, here and here.
- Joksas, D., Freitas, P., Chai, Z., Ng, W. H., Buckwell, M., Zhang, W. D., … & Mehonic, A. (2020). Committee machines—a universal method to deal with non-idealities in memristor-based neural networks. Nature Communications, 11, 4273. doi: 10.1038/s41467-020-18098-0.
- Mehonic, A., Joksas, D., Ng, W. H., Buckwell, M., & Kenyon, A. J. (2019). Simulation of Inference Accuracy Using Realistic RRAM Devices. Frontiers in neuroscience, 13, 593. doi: 10.3389/fnins.2019.00593.
BEng Electronic Engineering (2015-2018, UCL)
The modules that I found the most intellectually rewarding were Photonics and Communication Systems (ELEC215P), Control Systems I (ELEC3003), Numerical Methods (ELEC3030), Quantum Physics (PHAS2222), and Atomic and Molecular Physics (PHAS2224).
- Kenyon, A. J., Munde, M. S., Ng, W. H., Buckwell, M., Joksas, D., & Mehonic, A. (2019). The Interplay Between Structure and Function in Redox-Based Resistance Switching. Faraday discussions, 213, 151-163. doi: 10.1039/C8FD00118A.