Stephanie Miller

My hypothesis: Using only pre-snap data features, an algorithm can predict whether or not a QB will Pass the ball (versus Rushing it) better than the majority baseline. [Read More]

Correlation analysis allows me to verify whether or not one feature (or answer to the questionnaire) has a high similarity with another. It’s essentially a matrix that gives a 0 to 1 percentage that allows me to discern how important the relationship is — it can help me pinpoint which features are most likely to help me discern which questions were more likely to showcase a diagnosis, versus others. [Read More]

The role of diagnosis has changed dramatically for autism since it’s inclusion into mental health disorders in 1980. Prior to 1980, it should be noted, patients were often diagnosed with schizophrenia, and even then, only severe cases of autism were considered. At the time, the estimated ratio was 1 in 2400. As of 1980, there were 6 symptoms for autism, and the individual had to present all of the symptoms to a certain threshold in order to be diagnosed. [Read More]