Distributed data processing system and distributed data processing method

A distributed data processing system, comprising:

• a host computing device and at least one additional computing device having a processor, respectively executing a first machine learning model and at least one of a plurality of machine learning models, wherein the first machine learning model is larger than the plurality of machine learning models, wherein
• the host computing device comprises:
• a data processor, configured to process a received dataset, and comprising:
• a load balancer, configured to:
• select a first set of at least one computing device capable of executing the first machine learning model on the received dataset from the host computing device and the additional computing device to process samples of the dataset and generate a first prediction result by executing the first machine learning model;
• and
• select a second set of at least one computing device capable of executing each of the plurality of machine learning models on the received dataset from the host computing device and the additional computing device to process the samples of the dataset and generate at least a second prediction result and a third prediction result by executing the plurality of machine learning models;
• and
• a mode selector, configured to select a second machine learning model from the plurality of machine learning models based on a comparison of the first prediction result and the at least second prediction result and the third prediction result, wherein for the comparison, the first prediction result generated from the first machine-learning model for the samples of the dataset is utilized as a ground truth to evaluate the at least the second prediction result and the third prediction result generated for the same samples of the dataset, and
• the data processor is configured to instruct at least one of the second set of at least one computing device that is selected for executing the second machine learning model to process a rest part of the dataset by executing the second machine learning model.