Mastering the Stratified Approach for Machine Learning

<p>sMAPE</p> Signup and view all the answers

<p>Functions used to train and monitor the quality of a model</p> Signup and view all the answers

<p>Differentiability with respect to parameters</p> Signup and view all the answers

<p>To evaluate the performance of a model during training and testing</p> Signup and view all the answers

<p>They are calculated after the estimator is already created with use of different cost function</p> Signup and view all the answers

<p>To make decisions about probability cut-off points in a responsible and aware way</p> Signup and view all the answers

<p>To plot TPR and FPR for every probability cut-off point</p> Signup and view all the answers

<p>AUC ROC measures the tradeoff between true positive rate and false positive rate while AUC PR measures the tradeoff between precision and recall</p> Signup and view all the answers

<p>To evaluate the performance of a classification model</p> Signup and view all the answers

<p>To ensure that relative class frequencies are approximately preserved</p> Signup and view all the answers

<p>To select one best model on the testing sample</p> Signup and view all the answers

<p>A technique for evaluating a machine learning model and testing its performance</p> Signup and view all the answers

<p>Because it uses different portions of the data to validate and train the model on different iterations</p> Signup and view all the answers

<p>Area Under the Curve ROC (AUC ROC)</p> Signup and view all the answers

<p>0.5 to 1</p> Signup and view all the answers

<p>Imbalanced classification tasks</p> Signup and view all the answers

<p>To combine precision and recall in a single visualization</p> Signup and view all the answers

<p>The probability that a uniformly drawn random positive has a higher score than a uniformly drawn random negative</p> Signup and view all the answers

<p>To get one representative number for the whole model</p> Signup and view all the answers

<p>Bias is the difference between the expected prediction and the correct model, and variance is the variability of the model prediction for given data points</p> Signup and view all the answers

<p>The simpler the model, the higher the bias, and the more complex the model, the higher the variance</p> Signup and view all the answers

<p>A metric that generalizes the mean absolute error (MAE) to the case of probabilistic forecasts</p> Signup and view all the answers

<p>A metric that measures the correlation between predicted classes and ground truth in binary classification</p> Signup and view all the answers

<p>The proportion of actual negative observations that are incorrectly classified as positive</p> Signup and view all the answers

<p>A combination of precision and recall in one metric</p> Signup and view all the answers

<p>The proportion of actual positive observations that are correctly classified as positive</p> Signup and view all the answers

<p>The proportion of observations predicted as positive that are actually positive</p> Signup and view all the answers

<p>To assess the quality of our model in a quasi-objective way and to execute hyperparameter tuning safely</p> Signup and view all the answers

<p>Leave-p-out</p> Signup and view all the answers

<p>A parameter that controls the learning process and is not estimable</p> Signup and view all the answers

<p>K-folds</p> Signup and view all the answers

<p>To cater to the specifics of the data, business problem, dataset size, imbalance, and computing resources</p> Signup and view all the answers

<p>A plot of model learning performance over experience or time</p> Signup and view all the answers

<p>To give an idea of how well the model is generalizing</p> Signup and view all the answers

<p>The trade-off between overfitting and underfitting in a model</p> Signup and view all the answers

<p>To avoid overfitting the model to the training data</p> Signup and view all the answers

<p>A dataset with unequal distribution of classes</p> Signup and view all the answers