Existing evidence in Sports Related Concussions(SRC) is insufficient to determine the best combination of measures in the use of evaluation tools. Three Machine learning methodes and SRC prediction rules are dicussed.
This talk describes a hybrid machine learning model based on the combination of human/knowledge based domains and computer generated feature rankings to improve accuracy of diagnosing SRC. Four feature selection criteria were constructed to create the optimal model, which was run on both Google AutoML and Random Forest for validation. The results show that the hybrid model has the best performance in predicting resolution time for 14-day and 28-day thresholds, along with outperforming previous published work. This research has significant impact in the use of domains and symptom ranking with machine learning to increase diagnosis accuracy. The hybrid model’s success in the use of domains to increase the efficiency of model training may have practical applications in real time situations.
Speaker: Melody Yin, 8th grade student at the Harker School
Register at weblink to receive connection information
Contact:Website: Click to Visit
Save this Event:iCalendar
Windows Live Calendar