The Studdridge Advantage

Many real-world data sets possess features that are not adequately covered by academic research on large-dimensional covariance matrix estimation, such as:

  1. Time-series correlation across observations
  2. ARCH/GARCH effects
  3. Some very large eigenvalues
  4. Some eigenvalues very close to zero
  5. The case where the number of variables is equal to the sample size
  6. Need to estimate the correlation matrix instead of the covariance matrix
  7. Prior knowledge of patterns in the orientation of the eigenvectors of the covariance matrix
  8. Complex data
  9. Missing data
  10. Aggregation of class probability with case probability

and many other non-standard situations. Studdridge International delivers tailor-made solutions that fulfill the needs of the most demanding and sophisticated clients in the world.