Ahmed Frikha

- Applied research in deep learning sub-fields involving multiple data sources. - Successful application of the developed methods to multiple image, time-series and tabular datasets. - Implementation of the developed algorithms as reusable packages and filing of 10+ patents and inventions. - Automation of distributed multi-GPU training and hyperparameter tuning on the cloud. - Contribution to customer projects from different business units. - Interview, mentoring and thesis supervision of master students.

- Research in meta-learning for physics-informed neural networks to accelerate simulations. - Method: Few-shot physics-informed domain adaptation via meta-operator learning.

Title: A Meta-Learning-Based Approach for Few-Shot Anomaly Detection on Highly Unbalanced Datasets (Completed with the highest distinction: 1.0) - Development of an approach to the unexplored class-imbalanced few-shot learning problem - Validation on self-created synthetic time-series datasets and on an industrial dataset of sensor signals recorded during manufacturing

- Investigation of potential uses of new technologies to optimize controlling and finance processes - Implementation of virtual assistants (chatbots) using IBM Watson and Amazon AWS

Term Project: Processing and Generation of Transport Demand Data and Implementation of a Clustered Ride-Sharing Algorithm in an Agent-based Simulation - Big Data (mobility demand data) collection, processing, analysis, generation and visualization - Choice, modification and implementation of a dispatching algorithm to simulate different Ridesharing scenarios in an Autonomous Door-to-Door Mobility-on-Demand System

- Research on Industry 4.0 topics with a focus on production (IoT and 3D-Printing) - Improvement of a tool for the product development

- Creation and adjustment of pneumatic circuit diagrams for rail vehicles - Writing and rectifying technical documentation - Contribution to the management of delivery dates and deadlines

- Concept development of the portfolio planning regarding framework components - Selection of suitable multicriteria optimization algorithms - Implementation of a tool chain in Matlab, Java and an SQL database

- Development of a tool for the management of customer requirements - Planning and analysis of the integration of traceability into the production plants - Contribution to an internal audit in a production site in Slovakia

- Manufacturing-related activities: milling, drilling, soldering, welding - Helping with the maintenance of production machines

Machine Learning

Computational Intelligence, Real-Time-Systems, Systems Engineering, Modeling and Simulation (final grade: 1.3)

- Automotive Engineering - Business administration