.. _auto_dataset_nelder_mead_simplex_optimizer: *DATASET NELDER-MEAD SIMPLEX OPTIMIZER* ======================================= Finds the optimum of an objective defined by a dataset of sample points using the Nelder-Mead simplex algorithm. Supports a surrogate-based search (inverse-distance-weighted interpolation + Nelder-Mead, bounded to the input column ranges) or a simple best-row selection. Returns a single-row dataset with the optimal input values and the corresponding interpolated output value. When to use ----------- Tagged: ``best-row``, ``dataset-optimization``, ``derivative-free``, ``idw``, ``inverse-distance-weighting``, ``nelder-mead``, ``optimization``, ``simplex``. Inputs ------ .. list-table:: :header-rows: 1 :widths: 20 20 20 20 20 20 * - Label - ID - Type - Default - Required - Description * - Dataset - dataset - dataset - — - ✓ - Tabular dataset of sample points where each row contains numeric values for one or more design (input) columns and a numeric objective (output) column; must be non-empty. * - Input Columns - input_column_names - text - — - ✓ - One or more numeric column names from the dataset that define the design-variable search space for the Nelder-Mead optimizer; must match column headers exactly. * - Output Column - output_column_name - text - — - ✓ - Single numeric column name from the dataset that represents the objective function to be minimized or maximized; must match a column header exactly. * - Direction - direction - select - minimize - - Optimization direction: 'minimize' (default) to find the lowest output value or 'maximize' to find the highest; leave default for cost/error minimization tasks. * - Method - method - select - surrogate - - Search method: 'surrogate' (default) builds an IDW surrogate and runs Nelder-Mead bounded to the input column ranges; 'best_row' simply returns the dataset row with the best observed output value without running an optimizer. * - Max Iterations - max_iter - text - 800 - - Maximum number of Nelder-Mead iterations (integer, default 800); only used when method='surrogate'; increase for higher-dimensional or poorly-conditioned problems. * - Tolerance - tol - text - 1e-06 - - Relative convergence tolerance on the simplex objective values (float, default 1e-6); only used when method='surrogate'; decrease for tighter convergence at the cost of more iterations. * - IDW Power - idw_power - text - 2.0 - - Power exponent for inverse-distance weighting in the IDW surrogate (float, default 2.0, typical range 1–3); higher values make interpolation more local; only used when method='surrogate'. Outputs ------- .. list-table:: :header-rows: 1 :widths: 20 20 20 20 * - Label - ID - Type - Description * - Optimum - dataset - dataset - Single-row dataset containing the optimal values for each input (design) column and the corresponding objective output value, either interpolated by the IDW surrogate or taken directly from the best dataset row. * - Status - status - string - Human-readable summary string reporting the method used, optimization direction, number of iterations, starting row index, and the achieved optimum value; also surfaces validation error messages on failure. Disciplines ----------- - ai_ml.surrogate - data.dataset.transform - design_exploration.optimization Runnable example ---------------- A runnable example is registered for this worker. Open the example workflow on the d3VIEW canvas: `/api/workflow/example?id=dataset_nelder_mead_simplex_optimizer `_ .. raw:: html

Auto-generated from platform schema. Worker id: dataset_nelder_mead_simplex_optimizer. Schema hash: 4843354723ef. Hand-curated docs in workerexamples/ override this page when present.