Sentence examples for profile window from inspiring English sources

To estimate the desirable conditional drop probabilities, A-greedy maintains a fixed-size sliding profile window.

Every time the profile window is updated with new metadata, the analysis phase is triggered.

The profile window may be empty after removing its out-of-date data.

After that, the AQP technique removes the out-of-date data values from the profile window (line 25).

Employing that rationale, each collected measurement is first added to the profile window and then used for change detection.

The size of this profile window must have a minimal length before the AQP technique applies its analysis phase.

The analysis phase is then triggered as soon as the profile windows that were updated have enough data items.

A-greedy adopts N profile windows to collect cost measurements and, subsequently, to derive cost estimates for the input filters (Section 4.1.2 of [6]).

If during change detection we find changes only for a subset of those characteristics, we can flush only the out-of-date contents of the corresponding profile windows.

Regarding A-greedy (as the technique has been employed after considering seven profile windows of different sizes and, thus, for each query we got seven different solutions), we considered for each query the overall cost of each possible solution.

If we keep historic data relying on packets of the last two weeks, one week, or one day to collect statistics regarding the filter drop probabilities, then we build orderings with 30%% approximately higher per-tuple processing cost (than the ones found when employing smaller size profile windows) due to out-of-date statistics.