Data is what drives machine learning – yet it’s expensive to label and provision it for the purpose of training systems. Moreover, data quality and distribution are important factors to consider in maximizing the performance of powerful ML algorithms. Not only do intelligent data selection methods have to be developed and tailored to the target use case, but the capabilities of development environments, deployment and recording pipelines are of paramount importance to reach this goal.
Learn more about the steps that CARIAD is taking to increase the efficiency of fleet data collection, such as maximizing the information over data ratio.