Round 12: Tossup 19

Models that change this quantity often use the perplexity hyperparameter. (10[1])A common method of changing this quantity fails to be effective when applied to the “Swiss roll” set. (10[2])Similarity between data points is defined using a t-distribution in a nonlinear method for changing this (10[1])quantity that is the most common stochastic neighbor embedding. In a binary classification model, (10[1])maximally “shattered” sets define a form of this quantity named for (10[1])Vapnik and (10[2])Chervonenkis. (10[1])This quantity decreases in a method that iteratively produces (10[1])orthogonal basis vectors (-5[1])in the direction of maximum variance. This quantity can be reduced through PCA and LDA, (10[6])which (-5[1])prevents combinatorially (-5[1])explosive (10[1])sampling problems resulting from its “curse.” (10[2])For 10 points, the size of a model’s feature space is (10[1])what quantity that equals two for points on a planar grid? (10[1])■END■ (10[3])

Back to tossups

Buzzes

Summary