Continuously Learning of Neural Networks

Context

Modern vehicles no longer rely exclusively on on-board systems; their electrical/electronic (E/E) architecture is being extended to the cloud. This paradigm shift includes the migration of functions such as climate control to a centralized, cloud-based environment. Such advances allow models to be continuously updated by processing data in real time or in small increments rather than training on a static data set.

Goals

The cloud-based outsourcing of vehicle functions, such as heating, ventilation and air conditioning (HVAC) functions from the vehicle, opens up new possibilities for dynamic optimization and adaptation of such functions. Online learning is a promising implementation method for this functionality.

• Overview and research into suitable ML-based methods
• Implementation of the designed concepts
• Evaluation and comparison of the selected method using a test data set

Requirements

• Programming experience in Python or C++
• Basic knowledge of neural networks
• Motivation and interest in solving technical problems independently