2. In a speaker verification experiment, the inter-speaker and intra-speaker score (usually from a classifier, for example the scores may represent Euclidean Distance) distributions are respectively represented by the vectors A and B where A - 12 3 4 5 3 4 5 6 345 6 2 3 4 5] and B-17 8 979 8 5 7 9 8 6 7 8 9 6 7 8. A threshold estimation method that makes Gaussian assumptions about the data estimates u otherwise (a) Use this method to estimate the classification threshold (b) What would be the False Reject Rate and False Accept Rate at this threshold? 2. In a speaker verification experiment, the inter-speaker and intra-speaker score (usually from a classifier, for example the scores may represent Euclidean Distance) distributions are respectively represented by the vectors A and B where A - 12 3 4 5 3 4 5 6 345 6 2 3 4 5] and B-17 8 979 8 5 7 9 8 6 7 8 9 6 7 8. A threshold estimation method that makes Gaussian assumptions about the data estimates u otherwise (a) Use this method to estimate the classification threshold (b) What would be the False Reject Rate and False Accept Rate at this threshold