An open source software that allows for variable bandwidth kernel estimators. It is free and can be used by anyone.
If you're looking for a reliable way to classify and estimate probability density functions with ease, Adaptive Gaussian filtering might be the solution you've been seeking. This software library offers a simple and powerful implementation of variable bandwidth kernel estimators for classification, PDF estimation, and interpolation.
One of the standout features of Adaptive Gaussian filtering is its speed and accuracy. This library incorporates several innovations that make it one of the fastest and most accurate supervised classification algorithms on the market.
The algorithms included in this package are designed to make your life easier. For starters, the software matches kernel width to sample density quickly and accurately by leveraging the properties of the exponential function. Additionally, calculations are restricted to a set of k-nearest-neighbours using a binary tree, which can be found in n log k time.
Another great feature of Adaptive Gaussian filtering is its ability to generate a pre-trained model by searching for the class-borders with guaranteed, superlinear convergence. This saves you time and energy, so you can focus on other aspects of your work.
Finally, the software excels at extrapolating conditional probabilities to provide solid knowledge of estimate accuracy. This is a critical component in any software library designed to estimate probability density functions.
Overall, if you're searching for a reliable software library that can handle classification, PDF estimation, and interpolation with ease, Adaptive Gaussian filtering is an excellent choice. Its fast and accurate algorithms, innovative features, and ability to extrapolate conditional probabilities make it a tool you won't want to be without.
One of the standout features of Adaptive Gaussian filtering is its speed and accuracy. This library incorporates several innovations that make it one of the fastest and most accurate supervised classification algorithms on the market.
The algorithms included in this package are designed to make your life easier. For starters, the software matches kernel width to sample density quickly and accurately by leveraging the properties of the exponential function. Additionally, calculations are restricted to a set of k-nearest-neighbours using a binary tree, which can be found in n log k time.
Another great feature of Adaptive Gaussian filtering is its ability to generate a pre-trained model by searching for the class-borders with guaranteed, superlinear convergence. This saves you time and energy, so you can focus on other aspects of your work.
Finally, the software excels at extrapolating conditional probabilities to provide solid knowledge of estimate accuracy. This is a critical component in any software library designed to estimate probability density functions.
Overall, if you're searching for a reliable software library that can handle classification, PDF estimation, and interpolation with ease, Adaptive Gaussian filtering is an excellent choice. Its fast and accurate algorithms, innovative features, and ability to extrapolate conditional probabilities make it a tool you won't want to be without.
What's New
Version 0.92: N/A