For privacy protection of subjects in electroencephalogram (EEG)-based brain-computer interfaces (BCIs), using source-free domain adaptation (SFDA) for cross-subject recognition has proven to be highly effective. However, updating and storing a model trained on source subjects for each new subject can be inconvenient. This paper extends Euclidean alignment (EA) to propose adaptive Euclidean alignment (AEA), which learns a projection matrix to align the distribution of the target subject with the source subjects, thus eliminating domain drift issues and improving model classification performance of subject-independent BCIs. Combining the proposed AEA with various existing SFDA methods, such as SHOT, GSFDA, and NRC, this paper presents three new methods: AEA-SHOT, AEA-GSFDA, and AEA-NRC. In our experimental studies, these AEA-based SFDA methods were applied to four well-known deep learning models (i.e., EEGNet, Shallow ConvNet, Deep ConvNet, and MSFBCNN) on two motor imagery (MI) datasets, one event-related potential (ERP) dataset and one steady-state visual evoked potentials (SSVEP) dataset. The advanced cross-subject EEG classification performance demonstrates the efficacy of our proposed methods. For example, AEA-SHOT achieved the best average accuracy of 81.4% on the PhysioNet dataset.

O2O (online to offline) service recommendation is a typical context-aware service application, which needs to provide the most suitable services to customers in time according to their user profile and current context. By means of composing various data sources, different O2O service recommendation strategies can be customized, which may lead to great performance difference. Incorrect or non-real-time service recommendation would not work well and even cause negatives consequences. As a result, how to evaluate the performance of different O2O service recommendation strategies and select the most suitable one has become a key problem in the field. Due to the diversity and the variability of context events, as well as the economic, legal, and ethical impact, it is difficult or even impossible for traditional methods to realize comprehensive evaluation of various service strategies. Based on the background, this paper proposes a computational experiment-based evaluation method of O2O service recommendation strategies, which mainly consists of three parts: customization of O2O service strategies, modeling of experiment system, and execution of experiment evaluation. As a case study, the method was applied to Food O2O service. Three kinds of service strategies were compared respectively under two different market environments. Experiment results show that the proposed evaluation method is effective.