YSU ML Midterm

Minimum 1 hour

• 4 big problems, each has 4 or 5 sub problems

• 1st has 2 - need to justify

• TF question cover all slides

• need to discriminate partial and nabla

• 2nd - discussion about the role

• 3, 4 is similar to the assignment about Bayesian rule and calculating conditional probability

• There is node coding problem

Expectation - marginalization 하고 e = xp(x) 이용, sample mean사용

1. Introduction

• Probability Space
• Sample Space and event E is subset of
• field F (Event Space A) is set of E which are closed under intersection and combination
• requires to formally define the probability
• Probability P: F → [0, 1]

2. Linear Regression

Classification

Logistic Regression

Parameter Estimation

MLE

MAP

Generative Learning

Naive Bayes

SVM

Kernel