Fundamentals of Machine Learning - IAT 1

Machine learning is a branch of ________ that enables computers to learn from data.

Artificial Intelligence (AI)

________ learning uses statistical methods to understand relationships between variables and make predictions.

Statistical

The three major types of machine learning are supervised, unsupervised, and ________ learning.

Reinforcement

A learning model based on distances and spatial relationships is called a ________ model.

Geometric

A learning model that uses probability distributions is called a ________ model.

Probabilistic

________ models are commonly used for binary classification problems.

Logistic

GLM stands for ________.

Generalized Linear Models

In a GLM, the ________ function connects the linear predictor to the expected  value of the response variable

Link

The difference between an observed value and its predicted value is called a  ________

Residual

________ residuals are calculated directly as the difference between observed and fitted values

Raw

In supervised learning, the model learns from data containing input variables and a  known ________ variable

Target

A parametric machine learning method assumes that the model can be described  using a fixed number of ________

Parameters

A ________ model describes the relationship between a dependent variable and  one or more independent variables

Linear (or Regression)

In simple linear regression, there is ________ independent variable

One

The general equation of simple linear regression is Y = \beta_0 + \beta_X +  ________

e_i (Random error term)

In the linear regression equation, β0 represents the ________

Intercept (or Baseline value)

In linear regression, β1 represents the ________ of the regression line

Slope

The method commonly used to estimate regression parameters is called the  ________ method

Least Squares (or Maximum Likelihood Estimation)

The estimated regression coefficients are chosen to minimize the sum of ________ squared errors

Residuals (or Errors)

The coefficient of determination is represented by ________ and measures the  proportion of variation explained by the regression model

R²

________ residuals are based on the contribution of each observation to the deviance

Deviance

Over dispersion occurs when the observed variance is ________ than the variance  assumed by the statistical model

Greater

________ regression is used to model count data 

Poisson

In Poisson regression, the response variable generally represents the number of ________  occurring in a fixed interval

Events

A Poisson random variable has its mean and variance theoretically ________

Equal

A statistical test used to determine whether an individual regression coefficient is  significantly different from zero is the ________ test

t (or Wald)

The ________ value is used to determine the statistical significance of a regression  coefficient

p

Multiple linear regression involves ________ or more independent variables

Two

In multiple linear regression, the regression coefficients represent the effect of each  predictor while holding the other predictors ________

Constant (or Fixed)

One important assumption of multiple linear regression is the absence of severe ________  among the independent variables

Multicolinearity