DY0-001
Free CompTIA DataX Practice Test
12 exam-style questions with full explanations — no sign-up. Score yourself, then close your gaps with the full course.
Exam fee ~$5445 exam domainsLevel Advanced2 timed practice tests in the course
✅ Free practice — no sign-up📝 Real exam-style questions💡 Detailed explanations💸 30-day money-back via Udemy
A data scientist builds a churn prediction model achieving 99% AUC on validation but only 62% AUC in production. Investigation reveals that a categorical feature was encoded using the target variable's mean computed on the entire dataset before splitting into train and test sets. Which BEST describes the problem and the correct fix?
Computing target-based encoding on the full dataset before splitting lets test-set target information leak into training features, artificially inflating validation performance. The fix is to fit encodings only on training folds (e.g., within a cross-validation loop) and apply them to validation/test data.
A data scientist trains a decision tree that achieves 99% accuracy on training data but only 62% accuracy on a held-out test set. Which technique BEST addresses this issue while preserving the tree-based modeling approach?
Random Forest averages many bagged trees trained on bootstrap samples with random feature subsets, which reduces variance and overfitting while remaining tree-based.
A company deployed a fraud-detection model into production. Over three months, precision dropped from 92% to 74%, but monitoring shows the incoming feature distributions still closely match the training data distribution. What is the MOST likely explanation?
Concept drift occurs when the underlying relationship between inputs and the target variable changes (e.g., fraudsters adopting new tactics), causing performance decay even when input feature distributions remain stable. This matches the scenario exactly.
A data scientist wants to test whether customer preference for product category (A, B, C) is independent of the marketing channel used to reach the customer (email, social media, search ads). Both variables are categorical. Which test should be used?
The chi-squared test of independence is designed to determine whether two categorical variables are statistically associated by comparing observed frequencies in a contingency table to expected frequencies under independence.
An e-commerce platform launches a new product line with zero purchase history. The current recommendation engine relies on collaborative filtering built from a user-item interaction matrix, so the new products never appear in recommendations because there is no interaction data to leverage. Which approach BEST addresses this cold-start problem?
Hybrid recommenders solve cold-start by falling back to content-based similarity (item attributes) when interaction data is absent, while still exploiting collaborative signals once enough interactions accumulate. This is the standard, exam-tested fix for the item cold-start problem.
A fraud detection dataset contains 0.5% positive (fraud) cases and 99.5% negative cases. Which metric BEST evaluates model performance for this highly imbalanced classification problem?
With extreme class imbalance, Precision-Recall AUC focuses on performance for the rare positive class and is far more sensitive to changes in false positives/negatives among the minority class than ROC-AUC, which can look deceptively high due to the large number of true negatives.
A dataset with 10,000 features and only 500 samples needs to be reduced to two dimensions specifically to visually inspect natural groupings in a scatter plot. Which technique is BEST suited for this visualization task?
t-SNE is a nonlinear technique specifically designed to preserve local neighborhood structure, making it ideal for visualizing clusters in 2D.
A team wants to deploy a new recommendation model with minimal risk. They plan to route only 5% of live production traffic to the new model, monitor online business metrics against the incumbent model, and roll back quickly if problems appear, while 95% of users continue receiving results from the current model. Which deployment strategy is being described?
Canary deployment gradually exposes a small percentage of live traffic to a new model version to detect issues early and enables fast rollback, which is exactly the risk-mitigation pattern described.
A researcher wants to compare mean conversion rates across five different website designs (an A/B/C/D/E test) using continuous conversion rate data. Which approach best controls the family-wise error rate compared to running multiple pairwise tests?
One-way ANOVA first tests whether any of the five group means differ overall, and Tukey's HSD then performs pairwise comparisons while controlling the family-wise error rate, avoiding the inflated Type I error of running many uncorrected t-tests.
A support team wants to automatically extract specific product names, dates, and monetary amounts from incoming customer tickets so they can be routed to downstream billing and inventory systems. Which NLP technique is BEST suited for this task?
NER is specifically designed to identify and classify spans of text into predefined categories such as product names, dates, and monetary values, which is exactly the extraction task described.
Data is considered Missing Not At Random (MNAR) when:
MNAR occurs when the probability of a value being missing is related to the unobserved value itself — for example, individuals with very high or very low incomes being more likely to withhold that information. This is the hardest missingness mechanism to handle because it cannot be corrected using observed data alone.
A team must build a model to translate variable-length English sentences into French, capturing dependencies between words that may be far apart in the sequence. Which architecture is BEST suited for this task?
Transformers use self-attention to directly relate any two tokens in a sequence regardless of distance, effectively capturing long-range dependencies without the sequential bottleneck of RNNs.
Ready for the real thing?
The full course has two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed answer explanations.
Start my full course on Udemy →DY0-001 exam — quick answers
How much does the DY0-001 exam cost?
The exam fee is approximately $544 and varies by region — confirm current pricing with the certification vendor before you book.
What topics are on the exam?
It covers 5 domains: Modeling, Analysis, and Outcomes (24%), Machine Learning (24%), Operations and Processes (22%), Mathematics and Statistics (17%), Specialized Applications of Data Science (13%). The full course has a dedicated chapter, lab and practice-test coverage for each.
Is this practice test really free?
Yes — all questions on this page are free with explanations and no sign-up. The paid Udemy course adds two full-length timed exams, video lessons and hands-on labs.
Will this prepare me for the real exam?
The questions mirror the real exam's style and are mapped to the official domains. This is exam-focused preparation — combine the free test with the full course's timed simulations to gauge your readiness.