Crowdly

By eliminating uncertainty in environmental data

By revealing functional groups and ecological gradients relevant for management

By replacing ecological field observations

They eliminate uncertainty in monitoring systems

What best describes the role of classification model outputs in environmental risk management?

They provide exact forecasts of future environmental conditions

They replace expert judgment entirely

They support decision-making by identifying and prioritizing risk categories

Because environmental datasets are often highly imbalanced, with rare hazardous events

Why is overall accuracy often insufficient for evaluating classification models in environmental applications?

Because environmental data are always noisy

Because accuracy cannot be computed for multi-class problems

Because accuracy ignores model interpretability

Which algorithm is particularly well-suited for detecting oil spills or marine litter in satellite or drone imagery?

Logistic Regression

Support Vector Machines

Random Forests

Convolutional Neural Networks (CNNs)

Which evaluation metric is most critical when the goal is to minimize missed environmental hazards (false negatives)?

What is the main objective of applying classification models in environmental risk detection?

Reduce the dimensionality of environmental datasets

Group environmental observations into unlabeled clusters

Assign observations to predefined risk categories (e.g. hazardous vs. safe)

Predict continuous environmental variables such as temperature or pollutant concentration

Why is regression considered a supervised learning method?

Because it does not require labeled data

Because the model is trained using known input–output pairs

Because it only works on small datasets

Because the model learns without a target variable

Why are regression models useful in environmental sciences?

They eliminate uncertainty in observations

They help predict and interpret continuous environmental variables

They only work for laboratory data

They replace physical measurements

What is the main purpose of splitting a dataset into training and test sets?

To evaluate model performance on unseen data

To remove outliers

To increase model complexity

To reduce computation time

Which variable represents the target (dependent variable) in the applied use case of this lesson?

Electrical conductivity

Water temperature