C_PAII10_35 Online Practice Questions and Answers

The analytics are generated during a phase called the training phase, using a training dataset. Depending on the situation, the training dataset may be cut into three sub-sets:

Note: There are 3 correct answers to this question.

A. A Training sub-set.

B. A Validation sub-set.

C. The proper TCP port,

D. A Testing sub-set

Once the PAi application is installed correctly, the most common expected error to be encountered is that the system does not have succulent rows of data to successfully build a predictive model. For the PAi application to work correctly, you need a minimum number of rows of data to train your predictive model. The number of required rows will vary from scenario to scenario and is dependent on the model type and the data.

There are two main ways to handle this issue:

Note: There are 2 correct answers to this question.

A. If you have access to HANA logs, you will be able to see the full APL exception details and can investigate further.

B. The Predictive Factory server contains a group of services handled by the Server Intelligence Agent (SIA). The services can be stopped and started with the Start and Stop Predictive Factory commands.

C. You can verify the data via normal ABAP tools to see if succulent records are present. For example, by doing a data preview on the Predictive Scenario data source, taking into account any model version footer's, you can see how many records are returned.

D. Communication between the Predictive Factory server and clients uses HTTP. You can configure encrypted communication (HTTPS) using Secure Sockets Layer (SSL) or Transport Layer Security (TLS)

The Decision Tree Each node in the tree displays: Note: There are 4 correct answers to this question.

A. The name of the expanded variable, for example Marital-status.

B. The categories on which the node population has been filtered, for example {Married-AFspouse;Nevermarried}

C. The Population of the node

D. The ratio of Positive Target (for nominal targets) or the Target Mean (for continuous targets)

E. The Usage entries are sparser and report a more global view of the activity on the server. They consist of one line per event on the following events: user connection, user disconnection, training a model, applying a model, saving a model.

Two indicators specific to Automated Analytics allow you to evaluate the performance of a model. Note: There are 2 correct answers to this question.

A. The predictive power, which is the quality indicator of the model, Automated Analytics User Guides and Scenarios Modeling Concepts P U B L I C 27

B. The prediction confidence, which is the robustness indicator.

C. Through Java Web Start: in this mode, a Web server needs to be installed and configured on the machine hosting the SAP Predictive Analytics server. It will deliver, through web pages, the necessary content in order for the client computer to run the app

If in the PAi Fiori application, a predictive model is shown in status error, more information about the cause may be found in the associated application log. If the PAi configuration is incorrect, which following messages may appear in the application log:

Note: There are 2 correct answers to this question.

A. Error: Invalid Table Message For example: invalid table name: sap.hana.pai.Engine::Entities.GTMP.PredictiveScenario: line 1 col 72 an indication that the PAi application is not installed correctly. The solution is to reinstall the PAi DU.

B. Error: ,QsuFLHQt Privilege For example: LnsucLent privilege: Not authorized This means that the PAi HANA role The solution is to consult your database administrator regarding role assignment.

C. In Predictive Factory, you declare one or more modeling servers. (A modeling server is the machine that hosts the Automated Analytics server that will run the modeling tasks). Predictive Factory communicates with the Automated Analytics server using CORBA. All connections to data and model sources are handled via the modeling server.

D. Network administration is simplified because all network traffic from the client is directed to the server. This means only two TCP ports need to be opened for an Automated Analytics installation.

The structure for a continuous variable is defined by several intervals each made of : Note: There are 4 correct answers to this question.

A. a lower bound ([ ]) that can be either open or closed,

B. a minimum value (Minimum)

C. a maximum value (Minimum)

D. With standalone installers on each targeted client system

E. a higher bound ([ ]) that can be either open or closed.

Data modeling with time series is subdivided into four broadly defined stages: Note: There are 4 correct answers to this question.

A. Defining the Modeling Parameters

B. Generating and Validating the Model

C. Analyzing and Understanding the Analytical Results

D. Using a Generated Mode E. The Name Server

What is time Series?

Note: There are 2 correct answers to this question.

A. Identify and understand the phenomenon represented by your time series.

B. Forecast the evolution of time series in the short and medium term, that is, predict their future values.

C. Different system account, as specified in a "User-Mapping" file. This file will specify, for each authenticated user name the name of a system account to be used. This feature is available only on Linux

D. The system account of the authenticated user, this is the default with the system authentication.

A user has created several analyses with a chain in the Predict room. What data can the user leverage in a visualization in the Visualize room when these analyses are executed successfully?

Note: There are 1 correct answers to this question.

A. Data from one component from one analysis

B. Data from one component from all analyses

C. Data from all components from all analyses

D. Data from all components from one analysis

In the large majority of cases, a first degree polynomial is sufficient for generation of a relevant and robust model. Using a higher degree of polynomial does not always guarantee better results than those obtained with a first degree polynomial. In addition, the higher the degree of polynomial you select:

Note: There are 3 correct answers to this question.

A. The more time needed to generate the corresponding model

B. The more time needed to apply the model to new datasets

C. The harder it is to interpret the results of modeling

D. The server pass phrase (optional)