Computer Vision News - March 2020

2 Summary Code with Us 6 If you use the above code, you will be able to access the fit using the confusion matrix. The specificity and sensitivity are also calculated. Both terms are widely used in medicine and denote the true positive rate and specificity measures the proportion of actual negatives that are correctly identified as such (e.g., what’s the percentage of healthy people without condition that were correctly identified). We also get a plot of the Area Under Curve (AUC) which shows the performance of the model with one single number (although there are some issues which we can leave for you to further research). The usual numbers for it are as following: • 0.90 - 1.00 = excellent • 0.80 - 0.90 = good • 0.70 - 0.80 = fair • 0.60 - 0.70 = poor • 0.50 - 0.60 = fail If you run the previous model, you’ll see that it gives an AUC of 0.91 which means that our model has excellent prediction powers (sic)! You can also have a look at the graph. That’s great news and I am looking forward to exploring with you the internals of the model’s prediction mechanisms next month. All images are Public Domain from Wikipedia and Wikimedia Common