Loading demo links...
What to study for each A00-480 domain
Statistics and machine learning
This domain contrasts statistics and machine learning, including supervised and unsupervised learning and explanatory versus predictive goals. It also covers labeled data, features and targets, and why statistical ideas of sampling and uncertainty underpin machine learning practice.
Key skills
supervised vs unsupervised learning, classification vs regression, explanatory vs predictive modeling, features and labeled outcomes
Fundamental statistical concepts
This domain covers populations, samples, parameters and statistics, probability, and common distributions including the normal distribution. It also addresses sampling distributions and the Central Limit Theorem, confidence intervals, hypothesis tests, p-values, errors, power, and correlation.
Key skills
probability and distributions, sampling and Central Limit Theorem, confidence intervals, hypothesis testing, Type I and Type II errors, correlation
Explanatory modeling using linear regression
This domain uses linear regression for explanatory modeling, interpreting intercepts and slopes and fitting models by least squares. It also covers residuals and diagnostics, R-squared and adjusted R-squared, model assumptions, multicollinearity, dummy variables, and model selection with AIC and BIC.
Key skills
slope and intercept interpretation, least squares and residuals, R-squared, assumptions and diagnostics, multicollinearity, AIC and BIC
Predictive modeling using logistic regression
This domain builds predictive models for binary outcomes with logistic regression, covering the logit link, odds, odds ratios, and maximum likelihood estimation. It also addresses classification thresholds, confusion matrices, sensitivity and specificity, ROC curves and AUC, calibration, and model assessment.
Key skills
logit and odds ratios, predicted probabilities, thresholds and confusion matrix, sensitivity and specificity, ROC and AUC, calibration
Statistical foundations of machine learning
This domain covers the statistical foundations that govern how machine learning models generalize, including overfitting, underfitting, and the bias-variance tradeoff. It also addresses train, validation and test data, cross-validation, data leakage, and regularization concepts such as ridge and lasso penalties.
Key skills
overfitting and bias-variance tradeoff, cross-validation, train-test splitting, data leakage, ridge and lasso regularization