An improved robust algorithms for fisher discriminant model with high dimensional data

Area of Science:

• Statistical Analysis
• Machine Learning
• Data Science

Background:

• Traditional Fisher discriminant methods struggle with high-dimensional data and are sensitive to outliers.
• Outliers can significantly degrade the performance of standard discriminant analysis techniques.
• Robust statistical methods are needed for reliable analysis of complex datasets.

Purpose of the Study:

• To develop an improved robust Fisher discriminant method for high-dimensional data.
• To enhance the performance of Fisher discriminant analysis in the presence of outliers.
• To introduce a computationally stable and accurate method for outlier-prone datasets.

Main Methods:

• Integration of the Minimum Regularized Covariance Determinant (MRCD) algorithm into the Fisher discriminant framework.
• Development of the MRCD-Fisher discriminant model.
• Comparative experiments with existing robust discriminant methods.

Main Results:

• The MRCD-Fisher discriminant model demonstrated superior robustness and accuracy compared to other methods.
• The method effectively handles high-dimensional data where variables exceed observations.
• The MRCD-Fisher discriminant ensures high data cleanliness and computational stability.

Conclusions:

• The MRCD-Fisher discriminant offers a practical and reliable solution for analyzing high-dimensional, outlier-prone datasets.
• This method contributes significantly to the field of robust statistical analysis.
• The MRCD-Fisher discriminant is a valuable tool for complex data analysis challenges.