DOI: 10.1109/CISP-BMEI.2018.8633126

A dataset paper with which auditory info about people talking is collected.

Here are the state-of-the-art as of Laguarta 2021 on the dataset proposed.

CS229: instead of solving a problem, learn from data to find a model to solve the problem approximately.

CS109: machine learning is the act of using some input to come up with some prediction, where the model is parameterized via a bunch of parameters. Hence, parameter learning approaches is how machine learning works.

CS205L: training data + model to estimate new data points with minimal error; the parallel is a “knowledge based system” with interpolation

For multi-class classification, the macroaverage is the average of statistical values (prec, recc, etc.) after they have been computed for each seperate class.

The microaverage is the combination of a confusion matrix BEFORE statistical values are computed.

For a charge to do something in a magnetic field, it has to have velocity; nothing happens without movement.

So:

\begin{equation} \vec{F}_{M} = q \vec{v} \times \vec{B} \end{equation}

To calculate: magnitude: \(qvB \sin \theta\) + right hand rule.

Radius

You maybe asked to find the radius of the path the particle takes, so:

\begin{equation} \frac{v^{2}}{r} = a \end{equation}

So, the net force here is:

\begin{equation} qvB = Ma \end{equation}

So plug in and solve

DOI: 10.3389/fnagi.2021.623607

One-Liner

Trained a bimodal model on speech/text with GRU on speech and CNN-LSTM on text.

Novelty

• A post-2019 NLP paper that doesn’t use transformers! (so faster (they used CNN-LSTM) lighter easier)
• “Our work sheds light on why the accuracy of these models drops to 72.92% on the ADReSS dataset, whereas, they gave state of the art results on the DementiaBank dataset.”

Notable Methods

Bi-Modal audio and transcript processing vis a vi Shah 2021, but with a CNN-LSTM and GRU on the other side.