Test Prep CFA-Level-I Exam Prep
CFA® Level I Chartered Financial Analyst (Page 382 )

• Confirming the data has the required columns and data types
• Handling missing or duplicate records
• Converting data into a suitable format
• Selecting or engineering useful features
• Splitting it into training, validation, and test sets

1. Prepare the training data
2. Select an algorithm or configure AutoML
3. Train the model
4. Evaluate its performance
5. Tune hyperparameters if needed
6. Deploy the model

• 25 – Traditional SMTP server-to-server mail transfer; often blocked for client submission and not inherently secure.
• 110 – POP3 for receiving email; unencrypted by default.
• 143 – IMAP for accessing email; unencrypted by default.
• 587 – Authenticated email submission and commonly secured with TLS.

• Namespace: The namespace associated with the code where the checkpoint was set.
• Time: When the checkpoint was captured or executed.

• Exception: Exceptions are viewed through debug logs or execution results, not as the primary Checkpoints tab information.
• Debug Statement: System.debug() output belongs in the Logs tab. A checkpoint captures program state, such as variables and their values, at a specific point in execution.

• netstat -tuln can show listening TCP ports.
• If the web service is not listening, or is bound only to 127.0.0.1, remote browsers cannot connect even though the server responds to ping.

• traceroute: Tests the path to the server, but ping already confirms basic reachability.
• tcpdump: Useful for detailed packet analysis, but it is more appropriate after checking whether a service is listening.
• arp: Helps troubleshoot local IP-to-MAC resolution, not an application-layer web connection.

• Determine risk tolerance: Tolerance defines how much variation or exposure the organization is willing to accept. Scenarios may help apply tolerance, but they do not primarily determine it.
• Determine risk appetite: Risk appetite is established at an enterprise or governance level. Scenarios help compare risks with the appetite; they do not normally establish it.
• Develop risk responses: Once risks are identified and assessed, scenarios support response planning. However, response development comes later.

• Disabling a compromised account
• Blocking a malicious IP address
• Isolating an endpoint
• Creating or escalating an incident ticket

• Virus signatures: Primarily associated with antivirus or endpoint protection.
• Security agent deployment: Usually handled by endpoint-management or security-agent platforms.
• Vulnerability scanning: Typically performed by vulnerability-management tools.

• Extracted text and images
• Page numbers and bounding polygons for citation or grounding
• Structured tables, including tables that continue across multiple pages
• Document structure such as sections and figures

• Document Extraction: Extracts basic text and images, but does not provide the full layout and structural metadata required here.
• Azure Content Understanding in Foundry Tools: A broader content-analysis service, not the single built-in Azure AI Search skill expected by this question.
• GenAI Prompt: Can transform or interpret content but does not inherently produce reliable document geometry and page-level polygons.

• Deterrent control: The signs warning that people may be filmed can discourage unauthorized or malicious activity. They do not directly stop the action; they influence behavior.
• Detective control: CCTV records or monitors activity so security personnel can identify events or investigate them afterward.

• Preventive: Stops an incident from occurring, such as access control.
• Corrective: Restores systems after an incident.
• Directive: Instructs people to behave in a certain way, such as a security policy.
• Compensating: An alternative control used when the preferred control is not feasible.

• Required response and resolution times
• Availability or uptime targets
• Performance standards
• Support obligations
• Remedies or penalties for failing to meet the terms

• SOW (Statement of Work): Defines the project’s scope, tasks, deliverables, and schedule.
• MOA (Memorandum of Agreement): Describes agreed responsibilities between parties, often for a specific activity.
• MOU (Memorandum of Understanding): Records a mutual understanding or intention, usually less formal than a contract.

• On-premises resources, such as an existing database
• Cloud resources, such as Azure services or virtual machines
• Connectivity that allows the application to use both environments

• Public cloud: resources are entirely provided through a public cloud provider.
• Private cloud: resources are dedicated to one organization, typically on-premises or privately hosted.

• B: Prompt caching reduces cost and latency for repeated content, but cached content still occupies the context window. It does not solve context accumulation.
• C: Switching to a smaller model changes model capacity but does not address the growing context or remove irrelevant data.
• D: Keeping everything indefinitely increases token usage, risks context-window limits, and can distract the model with outdated information.

• B: Não é preciso implantar o modelo em um endpoint do Amazon SageMaker, pois o Bedrock já oferece a infraestrutura de inferência.
• C: O SageMaker Model Registry é usado para gerenciar modelos no SageMaker, não é requisito para modelos customizados no Bedrock.
• D: Não existe uma etapa geral de “conceder acesso” ao modelo customizado como requisito principal de uso. O ponto essencial é adquirir o Provisioned Throughput.

• Accept: Acknowledges the risk without taking additional action.
• Transfer: Shifts the risk or financial impact to another party, such as an insurer.
• Mitigate: Reduces the likelihood or impact through controls like patching, MFA, or segmentation.
• Avoid: Eliminates the activity that creates the risk.

• Smartphones and tablets
• Laptops and workstations
• Other mobile or endpoint computing devices

• Reduce the training data: Usually makes overfitting worse.
• Add hyperparameters: Hyperparameters already control training; simply adding them does not solve the problem. Proper tuning could help, but that is not what the option states.
• Increase training time: More training can worsen overfitting if the model is already memorizing the training data.

• Self-signed certificate: Causes a browser security warning, not incorrect name resolution.
• Nameserver record: A DNS problem would typically affect multiple users, not just one workstation.
• IP helper: Relays services such as DHCP across network segments; it does not resolve website names.

• Product recall: Premature before confirming a violation or assessing risk.
• Notify the regulatory authority: May be required later, especially if the discrepancy affects safety or compliance, but first investigate and verify the facts.
• Inform production: Not the initial regulatory assessment step; production may become involved if a packaging or manufacturing error is identified.

• Lead investigator: Coordinates the investigation and directs the response.
• Information security representative: Provides security expertise and helps assess the incident.
• Technical representative: Preserves system evidence, takes forensic backups, and supports technical analysis.
• Legal representative: Advises on legal, regulatory, privacy, and evidence-handling requirements.

• Who owns security risks and decisions
• Who is responsible for implementing controls
• Who reports security performance to management
• How security responsibilities align with business objectives

• B — Compliance with audit requirements: Governance may support compliance, but audit compliance is not its main objective.
• C — Monitoring security incidents: This is an operational security activity, not the central purpose of governance.
• D — Creation of tactical solutions: Governance establishes direction and oversight; tactical solutions belong to management and operational teams.

Box 1: No

• configure_azure_monitor() configures Azure Monitor and the OpenTelemetry pipeline.
• LangChain requires its own instrumentation or callback tracer, such as AzureAIOpenTelemetryTracer.
• You must pass the tracer to the LangChain execution configuration so chains, tools, prompts, and model calls become spans in Application Insights.

Box 2: Yes

text 
OTEL_SERVICE_NAME=customer-agent 
OTEL_SERVICE_NAME=invoice-agent

• JDBC
• SnowSQL
• Snowflake Web Interface (UI)
• ODBC
• Python connector

• Escape to host: A virtualization attack where code escapes a guest VM and accesses the host system. Nothing in the scenario mentions virtualization.
• Browser extension: Can redirect or manipulate web-browser traffic, but not normally database-server traffic.
• Valid accounts: Describes using stolen or compromised credentials, not redirecting network traffic.
• Process injection: Injecting malicious code into a legitimate database or networking process could allow an attacker to alter how that process handles outgoing traffic.

• Azure Cosmos DB is a globally distributed NoSQL database service. It stores and retrieves application data, but it does not provide the complete environment for building, testing, and deploying predictive analytics models.
• Azure Machine Learning is the appropriate service for this requirement. It supports creating experiments, training and testing models, and deploying them for predictions.
• Azure Machine Learning Studio provides the browser-based interface for these tasks.

• Measurement bias: Errors in how features or labels are measured, such as unreliable camera data or inconsistent labeling.
• Observer bias: A human evaluator’s subjective judgment influencing observations or labels.
• Confirmation bias: Favoring information that supports an existing belief.

• The ERSPAN source is the device or interface whose traffic is being copied.
• The ERSPAN destination is the device receiving the encapsulated copy and forwarding it to the traffic analyzer.
• The local monitor session number is locally significant. Therefore, the source and destination do not need matching session IDs such as 47 or 48.
• The erspan-id identifies the ERSPAN stream and must correspond between the source and destination configuration.

• B. Conditional access policies

• D. Additional authentication factors

• Disciplinary actions: Punish users but do not prevent credential-based attacks.
• More regular account audits: Useful for finding inactive or excessive accounts, but less direct for stopping unauthorized logins.
• Content filtering: Primarily controls web or email content, not authentication.
• Review of permissions: Reduces privilege but may not stop unauthorized account access.

• QuickVisualize is used to create and render a Power BI report from a DataFrame.
• The typical pattern is:

python 
PBI_visualize = QuickVisualize( 
    get_dataset_config(df), 
    auth=device_auth 
) 


PBI_visualize

• get_dataset_config(df) prepares the dataset configuration.
• auth=device_auth supplies authentication for the Power BI connection.
• Evaluating PBI_visualize renders the report in the notebook.

• Box 1: No
• Box 2: Yes
• Box 3: Yes

• Remote and global work: Access is based on the approved device, not a specific office or geographic location.
• Seamless login: Users may not need to repeatedly enter credentials on a recognized device.
• Limited equipment: It uses devices employees already have rather than requiring separate hardware.

• Geotagging: Restricts access based on location and may prevent legitimate remote work.
• Smart cards: Require issuing and managing additional physical equipment.
• Time-based logins: Limit access by time and do not provide a seamless authentication experience.

• A: CEO-sponsored formal policy — demonstrates executive support, but sponsorship at one point in time may not sustain commitment.
• B: Security awareness training — primarily improves employee behavior, not senior management commitment.
• C: Periodic alignment reviews — ongoing reviews show that security remains relevant, supports business priorities, and uses resources appropriately.
• D: Senior management signoff on the strategy — useful for governance and accountability, but signoff is generally a one-time event.

• Floating IP must be enabled on the load-balancing rule.
• Session persistence should normally remain None.
• The health probe determines which replica should receive traffic.
• Client-IP persistence is unnecessary and can cause connections from the same client to remain associated with a backend replica longer than intended.

• Q9 — Client IP session persistence: No
• Q10 — Floating IP: Yes

• Microsoft Foundry is an enterprise platform for developers, data scientists, and IT teams.
• It supports the full AI application lifecycle: selecting models, connecting enterprise data, developing, evaluating, deploying, monitoring, and scaling solutions.
• It can include visual tools and low-code experiences, but it is not primarily designed to enable business users to build complete generative AI solutions independently.

• B: You still need to select and configure an appropriate model.
• C: Business users may participate through visual tools, but they are not the primary target audience.
• D: Foundry includes low-code tools, but it is not fundamentally a low-code-only platform.

• Insert is used when you are creating a new record. It saves the record to the table and then leaves the form.
• Update is used when editing an existing record. It saves the modifications and then leaves the form.

1. System prompts can authorize users — No

1. System prompts can reduce TPM consumption — Yes, with qualification

1. System prompts guide generative AI behavior — Yes

apex 
for (Account updatedAccount : Trigger.new) { 
    AccountTriggerHelper.notinyxternalSystem(updatedAccount.Id); 
}

• Too many callouts: The callout occurs asynchronously in the future method, and each future transaction can perform its permitted callouts.
• Too many SOQL queries: The SOQL query is inside the future execution, not repeatedly in the trigger transaction.
• Callout from triggers not supported: Direct callouts from triggers are prohibited, but the trigger delegates the callout to asynchronous Apex, which is the correct pattern.

• Defender for Servers is the Defender for Cloud plan that protects Azure VMs and generates security alerts that can be reviewed in Defender for Cloud.
• Azure Diagnostics extension collects guest OS monitoring data, but it does not onboard VMs to Defender for Cloud or provide Defender security alerts.
• Data connectors are used for Sentinel integrations.
• Azure Storage Analytics and workflow automation do not provide VM protection.

• Sales revenue is affected because internal sales are recorded at the transfer price.
• Operating profit is affected because revenue changes while costs may remain unchanged.
• Return on investment (ROI) is affected because ROI depends on operating profit and the division’s investment.
• Cost of components purchased is not affected by the transfer price. These are costs incurred when buying components from external suppliers, not the internal price charged for the division’s output.

1. insert con; ? 1
2. Database.setSavepoint(); ? 2
3. update con; ? 3
4. Database.rollback(sp_admin); ? 4

• Compute: Not used; no virtual warehouse is required.
• Data Storage: Not scanned again.
• Metadata: May help Snowflake manage and validate cached results, but it is not the Snowflake architectural layer identified as serving the request.
• Cloud Services: Handles query coordination, metadata services, authentication, and result-cache access.

• Dataset file: Use Import mode. Excel files are imported into Power BI, and the two Sales tables can be combined with Power Query.
• Report file: Use a live connection to the published dataset. This keeps the report separate from the dataset and avoids importing the data again.

• Report file: DirectQuery/live connection
• Dataset file: Import

• COE: Identifies the Center of Excellence, such as Payroll or Benefits.
• Topic Detail: Classifies the HR service within its COE; it does not identify the catalog item.
• Template: Supplies default values or case configuration.
• Record Producer: Represents the catalog item that employees use to submit the request and create the case.

• The Azure DevOps (Boards only) connector is designed specifically for Azure DevOps Analytics views.
• It authenticates directly to Azure DevOps and lets you select an existing Analytics view.
• Filters defined in the Analytics view are applied by Azure DevOps before Power BI retrieves the data, reducing the amount of data transferred.

• Azure DevOps Server (Boards only) is for on-premises Azure DevOps Server.
• OData Feed can consume OData endpoints, but it is not the dedicated Azure DevOps Analytics-view connector.
• OData queries are not the intended dedicated connector for this requirement.

• Review administrative-role membership ? PIM access reviews
• Require justification for continued membership ? PIM can require justification when activating or maintaining privileged access
• Alerts about administrator assignment changes ? PIM alerts and notifications
• History of administrator activation ? PIM audit history
• See changes made to Azure resources ? PIM activation/audit information can be correlated with Azure Activity Logs

• Base table — the table containing common inherited fields
• Parent table — the table directly extended by the child
• Other tables in the hierarchy

• Reconnaissance: Gathering information about the target.
• Weaponization: Creating or combining a malicious payload with a delivery method.
• Delivery: Sending the payload, such as through phishing.
• Exploitation: Triggering a vulnerability or weakness to gain execution or access.

• Warehouse size: Larger warehouses use more credits per unit of time than smaller ones.
• Number of clusters: A multi-cluster warehouse can run multiple clusters, and each active cluster consumes credits. Therefore, increasing the maximum or active cluster count can increase consumption.

• Number of users: More users may cause more workload, but users themselves do not consume credits.
• Amount of data processed: Snowflake warehouses are not charged directly by the amount of data scanned. However, processing more data may cause the warehouse to run longer, indirectly increasing credits.