2018/November Braindump2go 70-774 Exam Dumps with PDF and VCE New Updated Today! Following are some new 70-774 Real Exam Questions:
1.|2018 Latest 70-774 Exam Dumps (PDF & VCE) 67Q&As Download:
https://www.braindump2go.com/70-774.html
2.|2018 Latest 70-774 Exam Questions & Answers Download:
https://drive.google.com/drive/folders/1WVXCup_qKNm0iitL4rKQ_hsgZd6M_dQD?usp=sharing
QUESTION 52
Note: This question is part of a series of questions that use the same scenario. For your convenience, the scenario is repeated in each question. Each question presents a different goal and answer choices, but the text of the scenario is exactly the same in each question in this series.
Start of repeated scenario
You plan to create a predictive analytics solution for credit risk assessment and fraud prediction in Azure Machine Learning. The Machine Learning workspace for the solution will be shared with other users in your organization. You will add assets to projects and conduct experiments in the workspace.
The experiments will be used for training models that will be published to provide scoring from web services.
The experiment for fraud prediction will use Machine Learning modules and APIs to train the models and will predict probabilities in an Apache Hadoop ecosystem.
End of repeated scenario.
You plan to share the Machine Learning workspace with the other users.
You are evaluating whether to assign the User role or the Owner role to several of the users.
Which three actions can be performed by the users who are assigned the User role? Each correct answer presents a complete solution.
NOTE: Each correct selection is worth one point.
A. Create, open, modify, and delete datasets.
B. Create, open, modify, and delete experiments.
C. Invite users to the workspace.
D. Delete users from the workspace.
E. Create, open, modify, and delete web services.
F. Access notebooks.
Answer: ABE
Explanation:
Once a Machine Learning workspace is created, you can invite users to your workspace and share access to your workspace and all of its experiments. We support two roles of users:
User – A workspace user can create, open, modify and delete datasets, experiments and web services in the workspace.
Owner – An owner can invite, remove, and list users with access to the workspace, in addition to what a user can do. He/she also have access to Notebooks.
https://github.com/anthonychu/azure-content/blob/master/articles/machine-learning/machine-learning-create-workspace.md
QUESTION 53
Note: This question is part of a series of questions that use the same scenario. For your convenience, the scenario is repeated in each question. Each question presents a different goal and answer choices, but the text of the scenario is exactly the same in each question in this series.
Start of repeated scenario
You plan to create a predictive analytics solution for credit risk assessment and fraud prediction in Azure Machine Learning. The Machine Learning workspace for the solution will be shared with other users in your organization. You will add assets to projects and conduct experiments in the workspace.
The experiments will be used for training models that will be published to provide scoring from web services.
The experiment for fraud prediction will use Machine Learning modules and APIs to train the models and will predict probabilities in an Apache Hadoop ecosystem.
End of repeated scenario.
The users will use different data sources that follow a standard format. The users will receive results in a standard format by using the fraud prediction web service. The results will be saved to a location specified by the users.
You need to provide the users with the ability to get results for different risk tolerances without affecting the calculation of the model.
Which three modules should be configured to use the Web Service Parameters? Each correct answer presents part of the solution.
NOTE: Each correct selection is worth one point.
A. Evaluate Model
B. Import Data
C. Select Columns in Dataset
D. Export Data
E. Time Series Anomaly Detection
Answer: ABD
Explanation:
A: From scenario: The experiment for fraud prediction will use Machine Learning modules and APIs to train the models and will predict probabilities in an Apache Hadoop ecosystem.
BD: An Azure Machine Learning web service is created by publishing an experiment that contains modules with configurable parameters. In some cases, you may want to change the module behavior while the web service is running. Web Service Parameters allow you to do this task.
A common example is setting up the Import Data module so that the user of the published web service can specify a different data source when the web service is accessed. Or configuring the Export Data module so that a different destination can be specified
https://docs.microsoft.com/en-us/azure/machine-learning/studio/web-service-parameters
QUESTION 54
You need to identify which columns are more predictive by using a statistical method. Which module should you use?
A. Filter Based Feature Selection
B. Principal Component Analysis
C. Group Data into Bins
D. Tune Model Hyperparameters
Answer: A
Explanation:
In machine learning and statistics, feature selection is the process of selecting a subset of relevant, useful features to use in building an analytical model. Feature selection helps narrow the field of data to the most valuable inputs. Narrowing the field of data helps reduce noise and improve training performance.
Feature selection is an important tool in machine learning. Machine Learning Studio provides multiple methods for performing feature selection. Choose a feature selection method based on the type of data that you have, and the requirements of the statistical technique that’s applied.
https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/feature-selection-modules
QUESTION 55
You are building a classification experiment in Azure Machine Learning.
You need to ensure that you can use the Evaluate Model module in the experiment.
Which two actions should you perform? Each correct answer presents part of the solution.
NOTE: Each correct selection is worth one point.
A. Connect the input of the Score Model modules to the output of the Evaluate Model module.
B. Connect the input of the Score Model modules to the output of the Train Model modules and the output Split Data modules.
C. Connect the output of the Score Model modules to the input of the Evaluate Model module.
D. Connect the output of the Score Model modules to the input of the Train Model modules and the input of the Split Data modules.
Answer: BC
Explanation:
Use the training data.
To evaluate a model, you must connect a dataset that contains a set of input columns and scores. If no other data is available, you can use your original dataset.
1. Connect the Scored dataset output of the Score Model to the input of Evaluate Model.
2. Click Evaluate Model module, and select Run selected to generate the evaluation scores.
Use testing data
A common scenario in machine learning is to separate your original data set into training and testing datasets, using the Split module, or the Partition and Sample module.
1. Connect the Scored dataset output of the Score Model to the input of Evaluate Model.
2. Connect the output of the Split Data module that contains the testing data to the right-hand input of Evaluate Model.
3. Click Evaluate Model module, and select Run selected to generate the evaluation scores.
https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/evaluate-model
QUESTION 56
You are working on an Azure Machine Learning experiment that uses four different logistic regression algorithms.
You are evaluating the algorithms based on the data in the following table.
Which model produces predictions that are the closest to the actual outcomes?
A. Model 1
B. Model 2
C. Model 3
D. Model 4
Answer: C
Explanation:
Mean absolute error (MAE) measures how close the predictions are to the actual outcomes; thus, a lower score is better.
Relative absolute error (RAE) is the relative absolute difference between expected and actual values; relative because the mean difference is divided by the arithmetic mean.
Root mean squared error (RMSE) creates a single value that summarizes the error in the model. By squaring the difference, the metric disregards the difference between over-prediction and under-prediction.
https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/evaluate-model
QUESTION 57
The manager of a call center reports that staffing the center is difficult because the number of calls is unpredictable.
You have historical data that contains information about the calls.
You need to build an Azure Machine Learning experiment to predict the number of total calls each hour.
Which model should you use?
A. Multiclass Logistic Regression
B. Boosted Decision Tree Regression
C. Decision Forest Regression
D. Poisson Regression
Answer: D
Explanation:
Poisson regression is intended for use in regression models that are used to predict numeric values, typically counts. Therefore, you should use this module to create your regression model only if the values you are trying to predict fit the following conditions:
The response variable has a Poisson distribution.
Counts cannot be negative. The method will fail outright if you attempt to use it with negative labels.
A Poisson distribution is a discrete distribution; therefore, it is not meaningful to use this method with non- whole numbers.
Incorrect Answers:
A: Logistic regression is a well-known method in statistics that is used to predict the probability of an outcome, and is particularly popular for classification tasks. The algorithm predicts the probability of occurrence of an event by fitting data to a logistic function.
B: Boosting means that each tree is dependent on prior trees. The algorithm learns by fitting the residual of the trees that preceded it. Thus, boosting in a decision tree ensemble tends to improve accuracy with some small risk of less coverage.
C: Decision trees are non-parametric models that perform a sequence of simple tests for each instance, traversing a binary tree data structure until a leaf node (decision) is reached.
https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/poisson-regression
QUESTION 58
You have an Apache Spark cluster in Azure HDInsight. The cluster includes 200 TB in five Apache Hive tables that have multiple foreign key relationships.
You have an Azure Machine Learning model that was built by using SPARK Accelerated Failure Time (AFT) Survival Regression Model (spark.survreg).
You need to prepare the Hive data into a single table as input for the Machine Learning model. The Hive data must be prepared in the least amount of time possible. What should you use to prepare the data?
A. a Hive user-defined function (UDF)
B. Spark SQL
C. the GPU
D. Java Mapreduce jobs
Answer: A
Explanation:
Create features for data in an HDInsight Hadoop cluster using Hive queries. Feature engineering attempts to increase the predictive power of learning algorithms by creating features from raw data that facilitate the learning process. You can run HiveQL queries from Azure ML, and access data processed in Hive and stored in blob storage, by using the Import Data module.
https://docs.microsoft.com/en-us/azure/hdinsight/hdinsight-machine-learning-overview#azure-machine-learning-and-hive
QUESTION 59
You are building an Azure Machine Learning workflow by using Azure Machine Learning Studio.
You create an Azure notebook that supports the Microsoft Cognitive Toolkit.
You need to ensure that the stochastic gradient descent (SGD) configuration maximizes the samples per second and supports parallel modeling that is managed by a parameter server.
Which SGD algorithm should you use?
A. DataParallelASGD
B. DataParallelSGD
C. BlockMomentumSGD
D. ModelAveragingSGD
Answer: A
Explanation:
Data-Parallel Training with Parameter Server
Parameter server is a widely used framework in distributed machine learning. The most important benefit it brings is the asynchronous parallel training with many workers.
You can use Data-Parallel ASGD.
Incorrect ANswers:
D: Model-Averaging SGD generally converges more slowly and to a worse optimum, compared to 1-bit SGD and Block-Momentum SGD, so it is no longer recommended.
https://docs.microsoft.com/en-us/cognitive-toolkit/multiple-gpus-and-machines
QUESTION 60
You plan to use Azure Machine Learning to develop a predictive model. You plan to include an Execute Python Script module.
What capability does the module provide?
A. importing Python modules from a ZIP file for execution in a Machine Learning experiment
B. performing interactive debugging of a Python script
C. saving the results of a Python script run in a Machine Learning environment to a local file
D. returning multiple data frames
Answer: A
Explanation:
A common use-case for many data scientists is to incorporate existing Python scripts into Azure ML experiments. Instead of requiring that all code be concatenated and pasted into a single script box, the Execute Python Script module accepts a zip file that contains Python modules at the third input port. The file is unzipped by the execution framework at runtime and the contents are added to the library path of the Python interpreter.
https://docs.microsoft.com/en-us/azure/machine-learning/studio/execute-python-scripts
!!!RECOMMEND!!!
1.|2018 Latest 70-774 Exam Dumps (PDF & VCE) 67Q&As Download:
https://www.braindump2go.com/70-774.html
2.|2018 Latest 70-774 Study Guide Video: