Machine Learning Basics
Regression
Clustering
Bias, Variance, Error
Advanced Topics in ML
100

When a response is not given.

What are unsupervised methods?

100

The null hypothesis for coefficients in linear and logstic regression.

What is Bi <> 0?

100

Is minimized during cluster assignments in K-means.

What is within-cluster variation?

100

Measures the quality of fit or accuracy in regression models.

What is the Mean Squared Error?

100

Primary purpose of a loss function in machine learning.

What is measuring the error between predicted and actual values?

200

The model is too simple to capture patterns.

What is bias?

200

The probability when b0 = 1, b1 = -2, and xi = 0.5.

What is the p(y=1|0.5) = 0.5?

200

Occurs after observations are assigned to a cluster in K-means.

What is calculating the centroid?

200

Inherent noise or variability in data that cannot be reduced.

What is irreducible error?

200

Role of the regularization parameter λ.

What is adjusting the strength of the regularization?

300

The reencoding of input data in a neural network.

What is a representation?

300

Minimized when estimating parameters in linear regression.

What is sum squares of the residual?

300

Results in more balanced clusters?

What is complete linkage?

300

Primary goal of using bagging in machine learning.

What is decreasing variance?

300

Smaller learning rates in gradient descent reduce the risk of?

What is avoiding overshooting minimum?

400

When Xi and Yi are observed.

What is supervised learning?

400
Are linear in terms of log(P/(1-P))?
What are the parameters of the logistic model?
400

Does not require the specification of the number of clusters.

What is hierarchical clustering?

400

Addresses correlated trees in bagging by introducing randomness in feature selection?

What is Random Forest?

400

A change in distribution between training and testing data?

What is covariate shift?

500

The theoretical floor of classifier performance.

What is the Bayes Optimal Classifier?

500

The ratio that defines R2.

What is the SSR/SST or 1 - SSE/SST?

500

Done to avoid local minima in k-means clustering.

What is running with multiple initializations?

500

Observed relationship between True Positive Rate and and the False Positive Rate.

What is a positive correlation?

500

A cause of sparsely populated space?

What is the Curse of Dimensionality?