kNN local linear estimation of the conditional density and mode for functional spatial high dimensional data

Traditionally, regression problems are examined using univariate characteristics, including the scale function, marginal density, regression error, and regression function. When the correlation between the response and the predictor is reasonably straightforward, these qualities are helpful and instructive. Given the predictor, the response's conditional density provides more specific information regarding the relationship. This study aims to examine a nonparametric estimator of a scalar response variable's function of a density and mode, given a functional variable when the data are spatially dependent. The estimator is then derived and established by combining the local linear and the k nearest neighbors methods. Next, the suggested estimator's uniform consistency in the number of neighbors (UNN) is proved. Finally, to demonstrate the efficacy and superiority of the acquired results, we applied our new estimator to simulated and real data and compared it to the existing competing estimator.

Automated summary

Traditionally, regression problems are examined using univariate characteristics, but this study aims to examine a nonparametric estimator of a scalar response variable's function of a density and mode, given a functional variable when the data are spatially dependent. The estimator is derived and established by combining the local linear and the k nearest neighbors methods, and its uniform consistency in the number of neighbors (UNN) is proved. The new estimator is then applied to simulated and real data to demonstrate its efficacy and superiority compared to existing competing estimators.