system does not work as well for one type/group of people compared to another

training data really does matter: it may make generalized predictions based on a majority/minority class.

Because IID characteristic of input data, the majority will be over represented

Here’s a PSPACE-COMPLETE language:

\begin{equation} \forall x_1, \exists x_2, \forall x_3 \dots \dots \phi\qty(x_1, x_2, \dots, x_{n}) = 1 \end{equation}

Plot each value of \(\alpha=\) (i.e. probability mass this much) between qmodel vs. qtrue; a best fit would be diagonal

• Viewing computational linguistics from the length across linear algebra and linear structure
• Quantum algorithms and the necessary infra were being developed; and in the 2010s programmable quantum computers became showing up

Quantum is done over the complexes, which makes the normal linguistics done with the reals more powerful.

• want to infer the probability distribution of words based on their letters

  • Linearity breaks down: letter combinations in not commutative; and P(letter C) + P(letter A) != P(letters CA)

• instead of encoding letters as one-hot vectors; we encode these letters with matrices: adds more dimensions