Mar 2024 – present
PhD candidate · Electrical, Electronics and Communications Engineering
MIRABILIS: optical feedback, self-mixing sensing, single-pixel imaging and reliable machine learning. Doctoral research in progress.
Background & experience
Başak Ersöz
PhD candidate · Politecnico di Torino
01 / Education
Mar 2024 – present
MIRABILIS: optical feedback, self-mixing sensing, single-pixel imaging and reliable machine learning. Doctoral research in progress.
2021 – 2023
Thesis: Local Marginal Price Forecasting Using Different Machine Learning Approaches. Data from MISO, ISO New England and PJM electricity markets.
2014 – 2019
Faculty of Electrical and Electronics Engineering, İstanbul, Türkiye.
02 / Research experience
Mar 2024 – present
Politecnico di Torino · MIRABILIS
Experimental optical sensing and computational imaging within the PRIN 2022 MIRABILIS project. Signal acquisition, speckle-pattern recording, sensing matrices, regime discovery, classification and image reconstruction. Collaboration across Politecnico di Torino, Università di Bari Aldo Moro and Politecnico di Bari.

Oct 2021 – Feb 2024
LSTM-based forecasting of locational marginal prices during volatile electricity-market conditions, alongside comparative studies of ensemble-learning methods for ECG analysis.

03 / Industry R&D
Mar 2021 – Oct 2022
Ford Motor Company & Ford Otosan
End-to-end integration, validation and sign-off of telematics control units and connected-vehicle features for Ford Transit and Ford of Europe, with connectivity data analysis and cross-team troubleshooting across software, electronics, manufacturing and suppliers. Remote vehicle features send a request from a phone through the cloud into the vehicle, which made both the value and the security requirements of connected data very concrete.

Mar 2020 – Mar 2021
Ford Motor Company & Ford Otosan
Development and release of multimedia, navigation, audio and connectivity systems for Ford Trucks FMax, including the move from Windows CE to Linux-based multimedia, HERE navigation integration and Digital Audio Broadcasting. A feature that looks simple to the driver depends on sensors, control units, CAN communication, software, suppliers and validation all working together.

Jun 2018 – Mar 2020
Sensor-based appliance concepts, PCB prototypes and Arduino-based embedded systems, including demonstrators prepared for IFA Germany. I integrated a pressure sensor into a vacuum cleaner so the machine could infer the floor type and adapt its suction behaviour instead of asking the user to choose, and explored conductive, capacitive, humidity, temperature and load sensing to estimate dryness and fabric behaviour in dryers. In both cases the point was to turn a sensor reading into a decision the product could take on its own.

04 / Methods & tools
MATLAB · Python · Simulink · SQL · C (basic) · Numerical modelling · Reproducible data processing · Scientific visualisation
PCA · t-SNE · DBSCAN · Feature engineering · Supervised and unsupervised learning · CNNs and U-Net · Autoencoders for reliability and anomaly scoring · LSTM / GRU for temporal states · Domain adaptation · Model selection · Per-class evaluation
Self-mixing interferometry · Semiconductor lasers · Speckle patterns · Single-pixel imaging · Compressed sensing · Sensing-matrix construction · Inverse reconstruction · SSIM and PSNR evaluation · Experimental acquisition and alignment
Sensor technologies · Embedded sensing · CAN · I²C · Bluetooth · Automotive connectivity · Requirements engineering · System integration and validation