The problem resides in the criteria to decide whether a gene is expressed or not, which depends on the algorithm used for summarization and normalization of data.
The resulting gene-level expression pattern changes depends on the set of probes used for summarization.
Some of the methods used for summarization can relate to user attention or saliency models [12, 21, 22], be domain-specific [23], i.e., be based on a specific topic, such as sports, news, documentary, movies, etc., or relate to the video's story [20], be based on users' preferences, the query context [18], or even focus on dominant concepts [24].
Assuming S samples drawn from the posterior, the samples are subsequently used for summarization, i.e. approximating the expected value of the parameters with Monte Carlo integration (Gelman et al., 2004), 𝔼 ≈ 1/ S∑ s =1 S θ(s ).
The time of pressing the buttons, the user who pressed the buttons, and the target utterance are recorded and used for summarization.
If a patient had more than one value within a given interval, the mean was used for summarization.
These types of metrics can be used for the summarization of any type of video and the quality measures may also be replaced with other metrics according to some specific input video types.
Principal components analysis (PCA) was used for graphical summarization and evaluation of the relationships among the studied samples.
Proportions, means, medians, frequency tables or cross-tabulations of important variables were used for data summarization and presentation.
RMA algorithm was used for data summarization (Bolstad et al. 2003) and quality control of samples was assessed by principle component analysis (PCA).
