Online Access Free DP-100 Exam Questions

No.# The correct answer is A. Yes.

Using the Last Observation Carried Forward (LOCF) method to impute missing data points does not affect the dimensionality of the feature set. LOCF works by filling in missing values with the most recent non-missing value from the same column. This approach ensures that the dataset retains the same number of rows and columns while providing a method to analyze the full dataset, including imputed values.

Key Points:
Dimensionality: The dimensionality of the dataset refers to the number of features (columns) and observations (rows). Since LOCF fills in missing values without adding or removing rows or columns, the dimensionality remains unchanged.
Analysis: By imputing the missing values, you can analyze the dataset without losing information or discarding any rows, allowing you to work with a complete dataset.
Therefore, the solution to use LOCF meets the goal of cleaning the missing values while preserving the dimensionality of the feature set.

No.# To generate summary statistics for a CSV file that includes the specified values (number of non-empty values, average, standard deviation, minimum, maximum, and percentiles), you should use Pandas because it provides built-in methods to handle CSV files and easily compute these statistics.

Here’s the Python code you would use:

import pandas as pd

# Load the CSV file
df = pd.read_csv('data/sample.csv')

# Generate summary statistics
summary = df.describe()

print(summary)
Explanation:
pandas: This library is essential for loading the CSV file and performing operations on the dataset.
describe: The describe() method in Pandas automatically computes the summary statistics for each numerical column, including count (non-empty values), mean (average), standard deviation, min, max, and the 25th, 50th (median), and 75th percentiles.
So, the correct selections are:

pandas (for loading the data and performing analysis)
describe (for generating the summary statistics)

No.# The correct answer is D. Use a Split Rows module to partition the data based on centroid distance.

The Split Rows module in Azure Machine Learning can be used to partition data based on specific criteria, such as the centroid distance. This technique is useful for tasks like clustering, where you want to segment the data based on distance to a central point (centroid), which could be relevant in determining a user's tendency to respond to an ad.

The other options are incorrect:

A and B: The Relative Expression Split module is used to split data based on conditional expressions, not on centroid distance or distance traveled.
C: While the Split Rows module is correct, partitioning based on "distance traveled to the event" isn't typically how you assess user tendency for ad response. The centroid distance is more applicable to clustering or proximity-based partitioning.

No.# The correct module to use is A. Split Data.

In Azure Machine Learning Studio, the Split Data module is used to divide data into two distinct datasets, which is often necessary for tasks like splitting the data into training and testing sets.

B. Load Trained Model is used for loading a previously trained model.
C. Assign Data to Clusters is used for assigning data points to clusters in unsupervised learning tasks like clustering.
D. Group Data into Bins is used for binning continuous data into discrete bins, not for splitting datasets.

No.# The correct answers are A. Uniform and C. LogUniform.

When tuning hyperparameters with Bayesian sampling in Azure Machine Learning, you can use the following parameter distributions for a learning rate:

Uniform: This distribution samples values uniformly from a specified range. It's useful when you want the learning rate to be selected randomly from a uniform interval.

LogUniform: This distribution samples values in such a way that the logarithm of the values is uniformly distributed. It is often used for parameters like learning rate, where the values can vary across several orders of magnitude.

Other options:

B. Normal and D. QNormal are used for parameters that follow a normal (Gaussian) distribution but are not typically ideal for a learning rate, which often benefits from a log-scale search.
E. Choice is used when you want to specify a discrete set of possible values, which is not ideal for continuous parameters like learning rate.

No.# The correct answer is C. Execute Python Script.

To add a new feature (column) such as CityName and populate it with a constant value (e.g., "London"), you would use the Execute Python Script module in Azure Machine Learning Studio. This module allows you to run custom Python code to manipulate your dataset, which is ideal for tasks like adding new features or columns programmatically.

For example, you could write a simple Python script that adds the "CityName" column to your dataset and sets all its values to "London."

Other options:

A. Edit Metadata is used to change the data type or labels of existing columns, but it cannot be used to add new columns.
B. Preprocess Text is used for text-specific feature engineering, such as tokenization, but it doesn't add new columns directly.
D. Latent Dirichlet Allocation (LDA) is for topic modeling in natural language processing, not for adding or manipulating dataset features.

No.# The correct answers are:

A. Registered dataset and E. Azure Blob storage container through a registered datastore.

Registered dataset: In Azure Machine Learning, you can import data into the pipeline by using datasets that have been registered with the service. Registered datasets are reusable and can be easily accessed across multiple pipelines and experiments.

Azure Blob storage container through a registered datastore: Azure Machine Learning supports importing data from Azure Blob Storage via a registered datastore. This allows seamless access to large volumes of data stored in Blob storage for training and experimentation.

Other options:

Azure SQL Database (B), URL via HTTP (C), and Azure Data Lake Storage Gen2 (D) are valid data sources for Azure services but are not directly applicable to the "Import Data" component in Azure Machine Learning Designer without some additional setup or integration through datastores.

At first, i was not sure about these DP-100 practice materials. I doubt it is up to date or not. But now with the certification, i can tell you it is the latest and valid.

I have successfully passed DP-100. Thank you freecram

This is a good DP-100 practice dump to use for preparing for the DP-100 exam. I passed the exam by the first try. Would recommend it to you!

No.# *Threshold optimizer technique
*Binary classification model

https://learn.microsoft.com/en-us/azure/machine-learning/concept-fairness-ml?view=azureml-api-2#mitigation-algorithms

No.# *Create a tabular dataset
*Create a compute cluster
*Create an experiment
*Create an automated ML job
Auto ML only Support tabular dataset

No.# Sales Data, Shop, [2017,2018]

No.# Split Data

No.# Create an instance of the OnlineDeploymentOperations class

No.# Command
Code

https://learn.microsoft.com/en-us/training/modules/run-training-script-command-job-azure-machine-learning/3-run-script-command-job

No.# The answer is
Shape
0

No.# Container Name
Wasps

No.# event grid
Function

No.# C
Azure Data brick