What is the correct definition of a hypothesis?
knowledge based on evidence and facts collected via repeatable tests to support a theory.
a proposed explanation made on the basis of limited evidence as a starting point for further investigation.
observational collection of signals and responses to stimuli in order to define a theory.
A hypothesis is a tentative, testable statement that proposes an explanation for an observed pattern or relationship. It is formed from limited evidence, prior knowledge, or reasoning, and it serves as the starting point for further investigation rather than a conclusion. In data science, a hypothesis might state that customers who receive a discount code are more likely to renew. It is expressed so that data can support or contradict it, usually as a null and an alternative hypothesis, and it is evaluated through experiments or statistical tests. Option B captures this definition. Option A describes a theory: an explanation supported by substantial evidence and repeatable testing, which has survived repeated attempts at refutation and therefore carries far greater confidence than a hypothesis. Option C describes observation, the systematic gathering of information about phenomena, which supplies the evidence from which hypotheses are drawn but is not itself a proposed explanation. The scientific method links the three: observation leads to a question, the question yields a hypothesis, testing generates evidence, and accumulated, consistent evidence may eventually support a theory. Distinguishing these terms is important because treating an untested hypothesis as established fact is a frequent source of analytical error. Reference: Data Science Fundamentals Study Guide – Data Science Concepts: The Scientific Method and Hypothesis Formulation.
===============
A model achieves near-perfect results on the training data but performs poorly when applied to new, unseen data. This behavior is MOST likely an example of:
Underfitting
Overfitting
Normalization
Overfitting arises when a model learns the training data too closely, absorbing random noise and idiosyncratic details in addition to the genuine underlying pattern. The model therefore achieves very low error on the data it was trained on but generalizes badly, because the quirks it memorized do not recur in new observations. The telltale sign is a wide gap between training performance and validation or test performance. Overfitting is encouraged by excessively complex models, too many features relative to the number of observations, and insufficient data. Remedies include collecting more data, simplifying the model, selecting fewer features, applying regularization, pruning decision trees, stopping training early, and evaluating with cross-validation. Underfitting, option A, is the opposite problem: a model too simple to capture the structure of the data performs poorly on both training and new data. Normalization, option C, is a preprocessing step that rescales feature values to a common range; it is not a model behavior and does not explain the pattern described. Understanding the bias and variance trade-off is central here, since overfit models exhibit high variance while underfit models exhibit high bias, and practitioners seek the balance that minimizes error on unseen data. Reference: Data Science Fundamentals Study Guide – Data Science Concepts: Model Evaluation, Overfitting and Underfitting.
===============
Development for Power BI is restricted to which of the following operating systems?
iOS
Unix
Windows
Power BI is Microsoft's business analytics platform, and its components serve different roles. Power BI Desktop is the authoring application in which developers connect to data, shape it in Power Query, build data models, write DAX measures, and design reports. It is a native application that runs only on Windows, so report development is restricted to that operating system; users on macOS or Linux typically rely on a Windows virtual machine or a remote desktop session. Option C is correct. The Power BI service, delivered through a web browser, is used to publish, share, and collaborate on reports and allows limited editing from any operating system, and mobile applications for iOS and Android allow consumption and interaction with published content. These tools can therefore be used on other platforms, but they do not provide the full development environment. iOS, option A, is a mobile platform for which only the viewing application exists, and Unix, option B, is not supported at all. This constraint affects tooling decisions in organizations whose analysts use Apple or Linux machines, and it is a reason some teams choose cross-platform tools such as Tableau for authoring. Candidates should distinguish the desktop, service, and mobile components when answering questions about platform support. Reference: Data Science Fundamentals Study Guide – Data Management: Data Analysis and Business Intelligence Tools (Power BI).
===============
Which of the following tasks is an example of a regression problem?
Determining whether an incoming email is spam
Predicting the sale price of a house from its size and location
Grouping customers into segments based on purchasing behavior
Supervised learning problems divide according to the type of output being predicted. Classification assigns an observation to a discrete category, whereas regression estimates a continuous numeric quantity. Predicting the sale price of a house from attributes such as floor area, number of rooms, and neighborhood yields a value that can fall anywhere along a numeric scale, so it is a regression task; linear regression, regression trees, and gradient boosting are common approaches, and performance is judged with error measures such as mean absolute error or root mean squared error. Option A, identifying spam, produces one of two labels and is therefore binary classification, assessed through metrics like precision and recall. Option C, grouping customers by behavior, has no predefined target variable at all; it is a clustering exercise belonging to unsupervised learning, in which the algorithm discovers structure on its own. A reliable shortcut is to ask what the model must output: a number on a continuum points to regression, a label from a fixed set points to classification, and the absence of any target points to unsupervised methods. Selecting the right problem type is the first step in choosing algorithms and evaluation criteria. Reference: Data Science Fundamentals Study Guide – Data Science Concepts: Machine Learning Approaches (Regression, Classification and Clustering).
===============
What is the MAIN purpose of a data warehouse?
hold data for later access.
serve as a documentation store.
facilitate the processing of inventories.
A data warehouse is a centralized, integrated repository that consolidates data from multiple operational and external sources and retains it, typically with history, so that it can be retrieved and analyzed afterward. It is organized around subjects such as sales, customers, or finance, and its data are loaded through ETL processes so that formats and definitions are consistent. The purpose is not to run day-to-day transactions but to hold that data for later access by analysts, reporting tools, and decision makers, which supports business intelligence, trend analysis, and performance management. Option A states this core role. Option B is incorrect because documentation stores and content management systems manage documents and files for collaboration and records purposes, and they follow different design principles. Option C describes an operational function. Inventory processing happens in transactional systems such as enterprise resource planning software, which are optimized for fast reads and writes of current records, whereas a warehouse is optimized for complex queries across large volumes of historical data. A warehouse might receive inventory data from those systems, but it does not perform the processing itself. This separation of operational and analytical workloads is a central reason warehouses are built. Reference: Data Science Fundamentals Study Guide – Data Management: Data Warehousing Concepts and Architecture.
===============
Which of the following is an example of exploratory analysis?
Confirming a preconceived assumption
Noticing patterns in visualization charts
Disproving a preconceived assumption
Exploratory analysis is an open-ended investigation of a data set carried out without a fixed hypothesis. The analyst examines distributions, plots relationships, and scans summaries to learn what the data contain, detect anomalies, and generate questions worth pursuing. Visualization is a core technique, since patterns such as clusters, trends, seasonality, and outliers are often visible in a chart long before they can be described numerically. Noticing patterns in visualization charts is therefore a characteristic exploratory activity, and option B is correct. Options A and C both begin from a preconceived assumption and use the data to test it, either supporting or refuting it. This is confirmatory, hypothesis-driven analysis, in which a claim is formulated in advance and evaluated through statistical tests with defined criteria. The two approaches are complementary. Exploration produces hypotheses and informs data preparation, and confirmation then tests them rigorously. A recognized caution is that patterns uncovered through exploration must not be validated on the same data that suggested them, because doing so overstates their significance; fresh data or a hold-out sample should be used. Knowing which mode of analysis is under way clarifies the standard of evidence expected and shapes how findings should be reported to stakeholders. Reference: Data Science Fundamentals Study Guide – Data Science Process: Exploratory and Confirmatory Data Analysis.
===============
Which company was responsible for creating the first autonomous database?
Oracle
SAP
IBM
Oracle announced the Oracle Autonomous Database in 2017 and made it generally available in 2018, positioning it as the first self-driving, self-securing, and self-repairing database offered as a cloud service. The product applies machine learning to tasks that traditionally required database administrators, including provisioning, performance tuning, indexing, patching and upgrading, backup and recovery, and threat detection. Because many of these operations occur automatically and online, the design aims to reduce human error, lower operating costs, and minimize planned downtime, allowing technical staff to concentrate on data modeling and application development rather than maintenance. It runs on Oracle's Exadata infrastructure and is offered in workload-specific forms for transaction processing and for data warehousing. Option A is therefore correct. SAP, option B, is best known for enterprise applications and for its in-memory database platform, SAP HANA, and IBM, option C, for Db2 and its broader data and AI portfolio; both vendors have added automation features since, yet neither introduced the first autonomous database. The concept is significant for data professionals because it illustrates the shift toward managed, AI-assisted data platforms in which routine administration is delegated to the system, changing the skills organizations prioritize. Reference: Data Science Fundamentals Study Guide – Data Management: Database Technologies and Emerging Trends (Autonomous Databases).
===============
Which data quality dimension is MOST directly violated when 15 percent of the records in a customer table have no value in the phone number field?
Accuracy
Completeness
Timeliness
Data quality is assessed against several dimensions, each of which isolates a different way data can fall short of its intended use. Completeness measures whether all required values are present, and it is breached whenever mandatory fields are empty or expected records are absent. A customer table in which a substantial share of rows lacks a phone number is a textbook completeness problem, because the information needed to contact those customers does not exist in the data set. It is typically quantified as the percentage of populated fields against those that should be populated and addressed through mandatory-field rules, validation at entry, or enrichment from other sources. Accuracy, option A, concerns whether recorded values correctly reflect reality; a phone number that is present but belongs to another person would violate accuracy, whereas a missing one does not. Timeliness, option C, concerns whether data are current enough for the purpose, so an outdated number that was once valid is a timeliness issue. Other recognized dimensions include consistency, validity, and uniqueness. Matching the symptom to the correct dimension allows an organization to select the proper remediation and monitoring control. Reference: Data Science Fundamentals Study Guide – Data Management: Data Quality Dimensions and Data Governance.
===============
In the Cross-Industry Standard Process for Data Mining (CRISP-DM), which phase comes first?
Data preparation
Business understanding
Modeling
CRISP-DM is a widely adopted, iterative methodology that structures data mining and data science projects into six phases: business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The sequence begins with business understanding because every later decision depends on a clear statement of what the organization is trying to achieve. During this phase the team defines the project objectives and success criteria from the business perspective, assesses available resources and constraints, identifies risks, and translates the business problem into an analytical question and a preliminary plan. Skipping or rushing it commonly produces technically sound models that do not address the real need. Data preparation, option A, belongs to the third phase and cannot be planned sensibly until the team has examined what data exist and how reliable they are. Modeling, option C, is the fourth phase and follows once suitable data sets have been constructed. Although the phases are listed in order, the cycle is not strictly linear; findings during data understanding or evaluation frequently send the team back to refine the business problem, and the process continues after deployment as conditions change. Reference: Data Science Fundamentals Study Guide – Data Science Process: Project Methodologies and the Data Science Life Cycle.
===============
What is the PRIMARY distinction between a data lake and a data warehouse?
Data lakes store internal data, while data warehouses store external data.
Data lakes are structured, while data warehouses are unstructured.
Data lakes store raw data, while data warehouses store scrubbed data.
The two repositories serve different analytical needs and differ chiefly in the state of the data they hold. A data lake ingests data as it arrives from source systems, without cleansing or remodeling, and retains it in raw form. Because the structure is imposed only when the data are queried, a lake is highly flexible and supports exploratory work and machine learning on diverse formats. A data warehouse holds data that have already been extracted, cleansed, conformed, and organized into a defined schema, commonly through an ETL process, so that business users can run consistent reporting and business intelligence queries with confidence. Option C expresses this contrast: raw versus scrubbed. Option A is wrong because both repositories can contain internal and external data; origin does not separate them. Option B inverts the actual relationship. Warehouses are structured and schema-on-write, while lakes accept structured, semi-structured, and unstructured content. The distinction carries consequences for users and cost. Lakes favor data scientists and engineers who need access to original detail, whereas warehouses favor analysts who need reliable curated figures. Many organizations now combine the two in a lakehouse design that layers warehouse-style management over lake storage. Reference: Data Science Fundamentals Study Guide – Data Management: Data Storage Architectures (Data Lakes and Data Warehouses).
===============
Approximately what percentage of values in a normal distribution fall within one standard deviation of the mean?
68 percent
95 percent
99.7 percent
The normal distribution is symmetric and bell-shaped, and its spread is described by the standard deviation. The empirical rule, sometimes called the 68-95-99.7 rule, gives the proportion of observations expected within fixed distances of the mean: approximately 68 percent lie within one standard deviation, approximately 95 percent within two, and approximately 99.7 percent within three. The correct response for one standard deviation is therefore 68 percent, option A. Options B and C are the figures associated with two and three standard deviations respectively, and a common error is to confuse the intervals. The rule has several practical uses. It supports quick judgments about whether an observation is unusual, since values beyond three standard deviations occur in well under one percent of cases and are often flagged as potential outliers. It also underlies z-scores, which express how many standard deviations a value lies from the mean, and confidence intervals built on the normal approximation. The percentages hold only when the data are approximately normal; heavily skewed or heavy-tailed data deviate from them, so analysts should check the distribution with a histogram or a normality test before applying the rule. Reference: Data Science Fundamentals Study Guide – Data Science Concepts: Probability Distributions and the Normal Distribution.
===============
What is the MAIN distinction between clustering and classification?
Clustering is unsupervised, while classification uses predetermined groupings.
Classification is performed as an automated real-time process, while clustering is performed manually.
Clustering is defined by analysts, while classification uses an algorithm.
The defining difference lies in whether the categories are known before the algorithm runs. Classification is a supervised technique: a model is trained on records that already carry labels, learns the features associated with each class, and then assigns new observations to one of those predetermined groups. Common algorithms include decision trees, logistic regression, and naive Bayes. Clustering is unsupervised. It receives unlabeled data and discovers natural groupings on its own by measuring similarity or distance between observations, as in k-means or hierarchical clustering; the resulting groups are interpreted afterward rather than specified in advance. Option A states this contrast accurately. Option B introduces a false dichotomy between real-time automation and manual processing; both methods are algorithmic, and either can operate in batch or streaming contexts. Option C reverses the roles: it is classification in which analysts or domain experts define the categories through labeling, whereas clustering relies on the algorithm to propose the structure, even though an analyst may choose parameters such as the number of clusters. Recognizing the supervised versus unsupervised distinction is essential for selecting an appropriate technique when labeled outcomes either exist or do not. Reference: Data Science Fundamentals Study Guide – Data Science Concepts: Machine Learning Approaches (Supervised and Unsupervised Learning).
===============
Which business analytics technique are business stakeholders using when they try to forecast holiday sales for the month of December?
Prescriptive analytics
Descriptive analytics
Predictive analytics
Business analytics is commonly organized into descriptive, diagnostic, predictive, and prescriptive techniques, and each answers a different question. Forecasting December holiday sales asks what is likely to happen in the future, which is the defining purpose of predictive analytics. The technique applies statistical models, time series methods, or machine learning to historical sales, seasonalit y, promotions, and economic indicators in order to estimate a future value together with its uncertainty. Descriptive analytics, option B, summarizes what has already occurred, such as last December's revenue by region, and produces reports and dashboards rather than projections. Prescriptive analytics, option A, goes a step further by recommending a course of action, for example how much stock to order or what discount to offer, typically using optimization or simulation built on top of a forecast. The scenario stops at estimating the sales figure itself, so no recommendation is involved. Stakeholders frequently move through the three in sequence: they review past results, forecast the coming period, and then decide on actions. Identifying which stage a described activity belongs to is a routine skill expected of data science practitioners working with business teams. Reference: Data Science Fundamentals Study Guide – Data Science Process: Types of Analytics (Descriptive, Predictive, Prescriptive).
===============
TESTED 21 Nov 2026
Copyright © 2014-2026 DumpsTool. All Rights Reserved