Two friends BSN (Bayesian) and FQN (frequentist) are trying to solve a K-class c

Question

Two friends BSN (Bayesian) and FQN (frequentist) are trying to solve a K-class clas- sification problem. Let Nk be the number of training samples in class Being a Bayesian, BSN did the following: » Estimated the class conditional densities using kernel density estimation technique with a Gaussian kernel (-1) Computed the prior probabilities as Pe = - . Performed classification using Bayes rule. Being a frequentist, FQN decided to follow the nearest neighbor approach. However, instead of simply using the standard one nearest neighbor rule, FQN decided to use all the training samples for estimating the class label of a test sample. In his approach every training sample votes for the class label of the test sample, and the strength of this vote decreases as the distance between the training and test samples increases Specifically, if Xir is a training sample with class label ytr and Xte is a test sample, then yr gets a vote as class label of Xte with strength exp class label that gets the strongest vote is assigned to Xte Who is going to perform better? Bayesian or Frequentist? Prove your answer mathematically.

Explanation / Answer

The top image is not clearly visible(kindly repost) and also could you attach additional information.

Related Questions

Navigate

• Browse (All)
• Browse T
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.