Fig 1.
Real-world time series with distinct periodic pattern.
Fig 2.
Overall structure of woa-wtconv-kanformer.
Fig 3.
Flowchart of whale optimization algorithm.
Table 1.
The Algorithm of WOA.
Fig 4.
Diagram illustrating the grouped convolution within the wtconv module.
Fig 5.
Illustration of wavelet deconvolution within the wtconv module.
Table 2.
The Algorithm of WTConv model.
Fig 6.
Linear layer of MLP network.
Fig 7.
Linear layers integrated with wav-kan.
Table 3.
Multivariate prediction results, model prediction length is S ∈ {96,192,336,720}, and fixed lookback length T = 96.
Fig 8.
Time series prediction plot with a forecast horizon of 96 time steps.
Fig 9.
Time series prediction plot with a forecast horizon of 336 time steps.
Table 4.
Comparison of model efficiency.
Table 5.
Performance comparison of each module.
Table 6.
Algorithm analysis of each module.
Table 7.
Multivariate prediction results, model prediction length is S ∈ {96,336}, and fixed lookback length T = 96.
Fig 10.
Time series prediction plot with a forecast horizon of 96 time steps.
Table 8.
Performance comparison of each model.
Table 9.
Multivariate prediction results, model prediction length is S ∈ {96,336}, and fixed lookback length T = 96.
Fig 11.
Time series prediction plot with a forecast horizon of 96 time steps.
Fig 12.
Time series prediction plot with a forecast horizon of 336 time steps.
Table 10.
WOA parameter optimization loss function value.
Table 11.
WOA parameter optimization time consumption.
Table 12.
The list of parameters with their respective definitions.
Table 13.
Parameter tuning iterative process.
Fig 13.
Iterative comparison of model loss function (mse) before and after woa optimization.
Fig 14.
Prior and following the utilization of the WOA module.
Table 14.
In a noisy environment multivariate prediction results, model prediction length is S ∈ {96,336}, and fixed lookback length T = 96.
Fig 15.
In a noisy environment multivariate prediction results, model prediction length is 96, and fixed lookback length T = 96.
Fig 16.
In a noisy environment multivariate prediction results,The global prediction graph with a prediction step size of 96.