ConditionNumberMethod
Enums · AiDotNet.Enums
Specifies different methods for calculating the condition number of a matrix.
For Beginners
The condition number is a measure of how sensitive a matrix is to changes in input.
Think of it like checking how stable a table is - a table with uneven legs (high condition number) will wobble a lot with small changes, while a stable table (low condition number) remains steady.
In machine learning and numerical computing:
- A low condition number (close to 1) indicates a “well-conditioned” matrix that produces reliable results
- A high condition number indicates an “ill-conditioned” matrix that might amplify small errors
- An infinite condition number means the matrix is singular (cannot be inverted)
This is important because ill-conditioned matrices can cause numerical problems in algorithms, leading to inaccurate results or slow convergence.
Fields
| Field | Summary |
|---|---|
ConditionNumberMethod InfinityNorm |
Calculates the condition number using the Infinity norm (maximum absolute row sum). |
ConditionNumberMethod L1Norm |
Calculates the condition number using the L1 norm (sum of absolute column values). |
ConditionNumberMethod PowerIteration |
Estimates the condition number using the power iteration method. |
ConditionNumberMethod SVD |
Calculates the condition number using Singular Value Decomposition. |
int value__ |