The Inaugural Lecture shows noble contributions in the areas of nonparametric statistics namely – Kernel Density Estimation (KDE) and its applications, Quality Control and recently, Data Science.
The choice of the bandwidth in KDE is examined using different methods for both the Univariate and Multivariate cases. This was done from the higher order derivatives approach, the hybrid approach and using boosting and bagging to reduce the two components of the error term (Asymptotic Mean Integrated Squared Error (AMISE) – Bias2 and the variance respectively.
New control charts were introduced in quality control for producers/manufacturers to maintain standards during the course of producing goods for daily human needs. These include the Bivariate control chart, Hotelling'sv T2 control limits and the permutation approach in obtaining control limits.
Finally, the application of KDE was shown in the areas of Agriculture, Material Science and Meteorology combining effectively with Data Science.