C1000-144

Converting a neural network into the newest version of TensorFlow or another deep-learning package is what type of performance drift or software decay?A . Data changesB . Concept driftC . Software changesD . Sampling bias and selection bias changesView AnswerAnswer: C

What are effective strategies for handling missing data? (Choose Two)A . Deleting all rows with any missing valuesB . Imputing missing values using statistical methodsC . Using a machine learning model to predict missing valuesD . Ignoring missing data during analysisView AnswerAnswer: BC

In the context of anomaly detection, what is the algorithm primarily searching for?A . Patterns that do not conform to expected behaviorB . The best way to group similar data pointsC . The optimal number of clusters in the dataD . The strongest predictors of a target variableView AnswerAnswer: A

Which type of plot would best illustrate the distribution of a single continuous variable?A . Line plotB . Bar chartC . HistogramD . ScatterplotView AnswerAnswer: C

In the context of classification, what does the term 'overfitting' refer to?A . The model performs equally well on the training and test datasetsB . The model performs poorly on both training and test datasetsC . The model performs too well on the training dataset but poorly on unseen dataD...

Which feature engineering technique can be used to simplify models and improve interpretability?A . One-hot encoding categorical variablesB . Normalizing continuous variablesC . Removing correlated featuresD . Increasing the number of featuresView AnswerAnswer: C

Which approach is recommended for prioritizing business opportunities when planning an MVP?A . Choosing the most straightforward implementation irrespective of impactB . Assessing the potential return on investment and strategic fitC . Prioritizing based on the preference of the project managerD . Focusing solely on technological innovationView AnswerAnswer: B

If the goal is to explore the central tendency and variability of a dataset, which types of plots would be most informative?A . Bar chart and line plotB . Histogram and box plotC . Scatterplot and heatmapD . Pie chart and line plotView AnswerAnswer: B

A Logistic Regression algorithm is used to classify images into four categories. If each image has a 5x5 pixel dimension, what is the the number of weights required (excluding biases) for this model?A . 200B . 100C . 300D . 400View AnswerAnswer: B

Principal Component Analysis (PCA) is a common technique for which of the following?A . RegressionB . ClassificationC . ClusteringD . Dimensional reductionView AnswerAnswer: D