Practise vocabulary for production ML monitoring: data drift, concept drift, PSI, model degradation, and alerting on distribution shifts.
0 / 10 completed
1 / 10
___ drift occurs when the statistical distribution of input features changes from the training distribution in production.
Data drift (covariate shift) is a change in the input feature distribution P(X). The model was trained on one distribution but receives inputs from a different distribution in production, degrading performance.
2 / 10
___ drift occurs when the relationship between inputs and outputs changes — the same inputs now map to different correct outputs.
Concept drift is a change in the target relationship P(Y|X). For example, a fraud model trained pre-COVID may see different fraud patterns post-COVID — same features, different ground-truth labels.
3 / 10
PSI (Population Stability Index) measures ___ between a reference distribution and a current distribution.
PSI quantifies how much a feature's distribution has shifted from the training baseline. PSI < 0.1 is stable, 0.1–0.25 is moderate shift, > 0.25 is significant drift requiring investigation.
4 / 10
A model ___ alert fires when a model's performance metrics (accuracy, AUC, F1) fall below a defined threshold in production.
Model degradation alerts detect when production performance drops — usually caused by drift, data quality issues, or a change in user behaviour. Detecting degradation early triggers investigation and potential retraining.
5 / 10
The process of automatically retraining a model when drift or degradation is detected is called ___ retraining.
Triggered (automated) retraining pipelines monitor drift and degradation metrics and automatically kick off a retraining run when thresholds are breached, keeping models fresh without manual intervention.
6 / 10
Reviewer: 'I'm seeing some high PSI scores on the user_age feature. The model was trained primarily on users aged 25-35, but now we're getting a lot of 60+ year olds. This could indicate concept drift.' Which of the following best describes the reviewer's concern?
This question tests understanding of concept drift. The reviewer highlights a shift in the *distribution* of features (user_age), not simply that the model's predictions are wrong. This shift can degrade performance over time because the model was built on assumptions no longer valid in production.
7 / 10
DevOps Engineer (to Data Scientist): 'Hey, I just ran a drift detection job and it flagged high performance degradation on the fraud_score model. The alert says 'Relationship Drift detected.' What does that mean specifically?
This question focuses on 'relationship drift'. It's not just about changes in feature distributions; it's about a change in *how* those features relate to the output. For example, previously a high transaction amount always indicated fraud, but now that's no longer true.
8 / 10
Developer (writing PR description for model monitoring): 'Implemented a new PSI threshold of 0.8 to trigger alerts. This aims to proactively detect concept drift in the customer_segmentation model. A score above this indicates a significant deviation from the training data distribution.'
This question tests understanding of PSI's purpose and interpretation. A score *above* a threshold indicates a significant deviation – it's a measure of difference between distributions, not a general metric of model quality.
9 / 10
Data Scientist (during daily standup): 'I'm working on setting up automated retraining for the recommendation engine. We're using drift detection to monitor feature distributions and performance metrics. If we see a significant shift, the system automatically retrains the model with the latest data.'
This scenario tests understanding of automated retraining. The key point here is that retraining happens *automatically* based on drift detection – it's a proactive measure to keep the model relevant and accurate as user behavior evolves.
This question focuses on interpreting an API response. The high PSI score (1.25) combined with the RMSE value indicates a significant deviation from the training data and suggests that performance degradation is likely due to concept drift in the 'user_purchases' feature.
What will I practise in "Model Monitoring & Drift Vocabulary"?
This module focuses on ML Model Serving — real workplace phrasing you'll use on the job. It contains 10 scenario-based multiple-choice questions with instant feedback.
Is this exercise free to use?
Yes. Every exercise on CoderSlingo, including this one, is free to use with no account or sign-up required.
How many questions does this exercise have?
This module includes 10 questions. Each one gives an immediate right/wrong result plus a full explanation of the correct phrasing.
What happens if I answer a question incorrectly?
You'll see the correct answer highlighted straight away, along with a plain-English explanation of why it's right and why the other options don't fit — mistakes are part of the learning here.
Can I retry the exercise if I want a better score?
Yes — use the 'Try again' button on the results screen to reset your score and go through the questions again. There's no limit on attempts.
Who is this ML Model Serving exercise for?
It's aimed at IT professionals with working English who want to sound more natural and precise around ml model serving — useful whether you're preparing for real conversations at work or just building confidence with the vocabulary.
Do I need an account to track my progress?
No account is needed. Your progress through the exercise is tracked locally in your browser for the current session, and you can replay the module at any time.
How is this different from reading a blog article?
This exercise is an interactive drill that tests and reinforces specific phrasing through multiple-choice questions with instant feedback, while blog articles explain concepts and vocabulary in prose. The two work well together.
Where can I find more ML Model Serving exercises?
See the ML Model Serving hub for more modules like this one, or browse the full Exercises page for other IT-English topics.
Can I complete this exercise on my phone?
Yes — every exercise on CoderSlingo is fully responsive and works on phones and tablets, so you can practise anywhere.