Sentence examples for mean trend segment from inspiring English sources

It is used to classify the mean trend segments and find those that are similar to the forecast mean trend segment.

The forecast mean trend segment (varvec{y}_{N- s-1 tau }^{(1)}) is projected to l buckets by the l hash table functions.

Unlike the models mentioned above, the proposed model regards the remainder, which is produced by SSA, as the fluctuation component, and LSH is also used to find similar segments to the forecast mean trend segment.

Due to the high randomness of the fluctuation component, similar segment searching is carried out on the mean trend segments, and a number of similar segments of the forecast mean trend segment are obtained by LSH. Figure 4 demonstrates an example of the result of similar segment searching.

Open image in new window Fig. 2 Framework of proposed forecast model Open image in new window Fig. 3 Reconstruction of the mean trend Open image in new window Fig. 4 Searching for similar segments to forecast mean trend segment Open image in new window Fig. 5 Results of 4-step ahead forecast of Elia data set.

As for the SSA-LSH-SVR model, the function of SSA in this model is the same as in the SSA-SVR model, and LSH is employed to find similar segments to the forecast mean trend segment, which are then used to form the training input to build an SVR model.

Both components are reconstructed in a phase space to obtain mean trend segments and fluctuation component segments.

The training input for SSA-SVR and SSA-LSH-SVR models are the mean trend segments only, but for the proposed model, it is the synthesis of similar mean trend segments and the corresponding fluctuation component segments.

Afterwards, as shown in Fig. 3, the mean trend will be reconstructed in the phase space and converted to the mean trend segments and fluctuation component segments, respectively.

Finally, in order to avoid error accumulation and fixed errors, the training input is the synthesis of the similar mean trend segments and the corresponding fluctuation component segments.

Finally, support vector regression is adopted for prediction, where the training input is the synthesis of the similar mean trend segments and the corresponding fluctuation component segments.