The Barkley Group, LLC
A) Artificial Intelligence Model Weaknesses
1) In the Data Space and Solution Spaces, the Tensorflow tool assumes the axes are at right angles to each other (ortho-normal), but for real problems this is often not the case. This means that the resulting AI solution are not as accurate as it could be. Using this non-orthonormal fact could lead to a better way of solving the AI equations and optimizing the results.
2) AI only works for numeric values. If you are trying to predict a solution where the training data set and test data set are only text files, not containing numeric values, then you must make approximations which reduce model accuracy.