This node is located in the Systems > Algorithms > Sampling folder in the Modules pane.
The Bayesian Adaptive Sequential Sampling (BASS) method is an iterative sampling algorithm based on the Bayesian optimization method. It uses the probabilistic properties of the DIM-GP metamodel to select optimal new designs based on the largest model uncertainty. The algorithm iteratively minimizes the model uncertainty with each new design. This method is particularly suitable for using adaptive sampling to train a globally accurate surrogate model with as few samples as possible. BASS is particularly useful when the generation of samples is very expensive. It can deal with noisy as well as failed designs. The algorithm is not limited to the number of samples or the number of input parameters and is applicable to regression and classification tasks for scalar output variables. It is possible to handle both discrete and continuous input parameters.
Node Settings
To access the options shown in the following table, double-click the Bayesian Adaptive Sequential Sampling (BASS) system on the Scenery pane and switch to the Settings tab.
| Option | Default Value | Description |
|---|---|---|
| Algorithm Options | ||
| Maximum iterations | 100 |
Maximum number of iterations. This value determines the number of calculations of the process chain, along with the iteration size. Maximum number of designs = number of start designs + maximum iterations × iteration size unless other termination criteria is met. |
| Iteration step size | 1 | Number of samples per iteration. This value determines the number of calculations of the process chain, along with the maximum number of iterations. |
| Number of new start designs | 20 | Designs to create in addition to the start designs for the initial model. |
| Stagnation iterations | 0 (no stagnation) | Maximum number of allowed iterations without improving the model quality measure PAM (similar to R2 ). |
| Target prognosis quality | 1 | Minimum prognosis quality metric to be achieved, similar to CoP. Algorithm terminates early if this criterion is fulfilled. |
| Job Submit Pattern (beta) | <jobscript> <arg1> <arg2> <arg3> |
Specifies the job submit pattern for bash or command prompt for
submitting jobs. Will be cast with If the last entry does not exist, the action will be skipped and if empty, arguments will be submitted without an optional argument command. |
| Training Options | ||
| Maximum epochs | 500 | Maximum number of training epochs used. Algorithm messages provide hints for the required number of epochs. |
| Batch size | 0 (disabled) | Activates batch training, if the value is non-zero and smaller than the number of samples. Recommended at higher sample sizes. |
| Noisy data | Off (disabled) | Enable this setting if the training data has noise. |
Additional Options
To access the options shown in the following table, in any tab, click .
| Option | Description |
|---|---|
| Limit maximum in parallel |
Controls the resource usage of nodes in the system. When the check box is cleared (default), a value is chosen to ensure the best possible utilization of the child nodes. When the check box is selected, set the value manually to specify how many designs are sent to child nodes, limiting the maximum degree of parallelism for all children. Ansys recommends keeping the check box clear. |
| Auto-save behavior |
Select one of the following options:
The project, including the database, is auto-saved (depending on defined interval) after calculating this node/system (either when the calculation succeeds or fails). By default, all parametric and algorithm systems have selected, all other nodes have selected. |