NanoSeeX Algorithm Patent: Multivariable Physical Model Fitting & Analysis
NanoSeeX has developed a proprietary algorithm capable of handling multivariable physical model fitting and analysis. This advanced technique utilizes data from multiple measurement systems (e.g., transmission, reflection, diffraction, scattering spectra). Each distinct physical mechanism is analyzed using our dedicated electromagnetic wave calculation engine (EM engine), which derives higher-order signals to determine individual COST Х values and the associated structure parameter variances (Δp1, Δp2,…, Δpn). This results in a comprehensive set of n COST Х values and corresponding variances (Δp1, Δp2,…, Δpn).
Next, we employ statistical methods to compute an optimal range for fitting the structure parameters. Within this fitting range, a randomized process generates various combinations of parameters, followed by cross feedback interactions, ultimately leading to an enhanced structural solution.
TW 113115628
The algorithm iteratively calculates and assesses whether the variance has decreased. If not, the process of parameter arrangement and cross feedback is repeated to achieve an improved structural solution.
Finally, we ensure that the resultant COST and FITTING VARIANCE are significantly reduced compared to the pre-loop values, guaranteeing more accurate and reliable results.