Question

Assume we have a two-class learning problem where the labeled data consists of 1,000 data points from one class, and 10 data points from the other class. A grad student is working on a machine learning algorithm and manages to get 99% accuracy using ten fold cross validation. Recall that 10-fold cross validation splits the data into 10 equal-sized chunks. For each chunk, the algorithm is trained on the other 9 and then tested on the held out chunk. The accuracies on each of the held-out chunks are then averaged to get an overall accuracy rate. Is this a big achievement for the algorithm?Justify your answer by describing why it should be difficuilt to get this accuracy on this kind of data, or giving a simple classification strategy whose accuracy is at least as good.

Answer 1

I don't know

Step-by-step explanation:

please just thank me I need the points