LVC 3 - Glossary of Notations
𝑡ℎ
Xi = Vector containing values of input features corresponding to the 𝑖 record, where 𝑖 ranges
from 1 to 𝑛
𝑡ℎ
Yi = Value of the output variable corresponding to the 𝑖 record
𝑌 = The predicted value of the output variable
𝐸 = Expected value or average
ϕ(𝑋) = A transformed version of feature vector 𝑋
ℎ(𝑋) = Non-linear classifier function
𝑇
θ = Transpose of vector θ
µ𝑘 = Mean vector for class 𝑘
𝐶𝑘 = Covariance matrix for class 𝑘
π𝑘 = Probability of a data point belonging to class 𝑘. These are called prior probabilities
𝑁(µ𝑘, 𝐶𝑘) = Normal distribution with mean µ𝑘 and covariance 𝐶𝑘
γ = Normalizing constant for class k in the normal distribution equation
𝑘
[email protected]R8L0PN473F
𝑃(𝑌 = 𝑘 | 𝑋) = Probability of data point belonging to class 𝑘 given the input features 𝑋. These are
called posterior probabilities
𝑃(𝑋 | 𝑌 = 𝑘) = Probability of 𝑋 given the output class 𝑌 = 𝑘
𝐶 = When covariances of all the classes are the same
𝐶𝑑𝑒𝑓 = The cost of someone being a defaulter
𝐶𝑙𝑜𝑠𝑡 = The cost of losing a customer
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒𝑠 𝑜𝑓 𝑐𝑙𝑎𝑠𝑠 𝑘
π𝑘 = 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒𝑠
= Estimate of π𝑘
µ𝑘 = Estimate of µ𝑘
𝐶𝑘 = Estimate of 𝐶𝑘
𝐿(𝑑𝑎𝑡𝑎; θ) = Likelihood function of the observed data
γ = Regularization hyperparameter
w = Weight in the likelihood equation for unbalanced data
||𝑥|| = Distance of a point 𝑥 from the origin
𝑤 = Weight in the weighted distance metric
This file is meant for personal use by
[email protected] only.
Sharing or publishing the contents in part or full is liable for legal action.
