Microsoft DP-100 Designing and Implementing a Data Science Solution on Azure Exam Practice Test

You need to implement an early stopping criterion on models that provides savings without terminating promising jobs.

Truncation selection cancels a given percentage of lowest performing runs at each evaluation interval. Runs are compared based on their performance on the primary metric and the lowest X% are terminated.

Example:

from azureml.train.hyperdrive import TruncationSelectionPolicy

early_termination_policy = TruncationSelectionPolicy(evaluation_interval=1, truncation_percentage=20, delay_evaluation=5)

Create Scatterplot

Summarize Data

Clip Values

You can use the Clip Values module in Azure Machine Learning Studio, to identify and optionally replace data values that are above or below a specified threshold. This is useful when you want to remove outliers or replace them with a mean, a constant, or other substitute value.

References:

https://blogs.msdn.microsoft.com/azuredev/2017/05/27/data-cleansing-tools-in-azure-machine-learning/

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/clip-values

Step 1: Sweep Clustering

Start by using the "Tune Model Hyperparameters" module to select the best sets of parameters for each of the models we're considering.

One of the interesting things about the "Tune Model Hyperparameters" module is that it not only outputs the results from the Tuning, it also outputs the Trained Model.

Step 2: Train Model

Step 3: Evaluate Model

Scenario: You need to provide the test results to the Fabrikam Residences team. You create data visualizations to aid in presenting the results.

You must produce a Receiver Operating Characteristic (ROC) curve to conduct a diagnostic test evaluation of the model. You need to select appropriate methods for producing the ROC curve in Azure Machine Learning Studio to compare the Two-Class Decision Forest and the Two-Class Decision Jungle modules with one another.

References:

http://breaking-bi.blogspot.com/2017/01/azure-machine-learning-model-evaluation.html

Box 1: Floating point

Need floating point for Median values.

Scenario: An initial investigation shows that the datasets are identical in structure apart from the MedianValue column. The smaller Paris dataset contains the MedianValue in text format, whereas the larger London dataset contains the MedianValue in numerical format.

Box 2: Unchanged

Note: Select the Categorical option to specify that the values in the selected columns should be treated as categories.

For example, you might have a column that contains the numbers 0,1 and 2, but know that the numbers actually mean "Smoker", "Non smoker" and "Unknown". In that case, by flagging the column as categorical you can ensure that the values are not used in numeric calculations, only to group data.

Scenario: Testing

You must produce multiple partitions of a dataset based on sampling using the Partition and Sample module in Azure Machine Learning Studio.

Box 1: Assign to folds

Use Assign to folds option when you want to divide the dataset into subsets of the data. This option is also useful when you want to create a custom number of folds for cross-validation, or to split rows into several groups.

Not Head: Use Head mode to get only the first n rows. This option is useful if you want to test a pipeline on a small number of rows, and don't need the data to be balanced or sampled in any way.

Not Sampling: The Sampling option supports simple random sampling or stratified random sampling. This is useful if you want to create a smaller representative sample dataset for testing.

Box 2: Partition evenly

Specify the partitioner method: Indicate how you want data to be apportioned to each partition, using these options:

Partition evenly: Use this option to place an equal number of rows in each partition. To specify the number of output partitions, type a whole number in the Specify number of folds to split evenly into text box.

Spearman's rank correlation coefficient assesses how well the relationship between two variables can be described using a monotonic function.

Note: Both Spearman's and Kendall's can be formulated as special cases of a more general correlation coefficient, and they are both appropriate in this scenario.

Scenario: The MedianValue and AvgRoomsInHouse columns both hold data in numeric format. You need to select a feature selection algorithm to analyze the relationship between the two columns in more detail.

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/feature-selection-modules

Step 1: Augment the data

Scenario: Columns in each dataset contain missing and null values. The datasets also contain many outliers.

Step 2: Add the Bayesian Linear Regression module.

Scenario: You produce a regression model to predict property prices by using the Linear Regression and Bayesian Linear Regression modules.

Step 3: Configure the regularization weight.

Regularization typically is used to avoid overfitting. For example, in L2 regularization weight, type the value to use as the weight for L2 regularization. We recommend that you use a non-zero value to avoid overfitting.

Scenario:

Model fit: The model shows signs of overfitting. You need to produce a more refined regression model that reduces the overfitting.

Box 1: Accuracy

Scenario: You want to configure hyperparameters in the model learning process to speed the learning phase by using hyperparameters. In addition, this configuration should cancel the lowest performing runs at each evaluation interval, thereby directing effort and resources towards models that are more likely to be successful.

Box 2: R-Squared

Box 1: 500

For Random seed, type a value to use as seed for randomization. If you specify 0 (the default), a number is generated based on the system clock.

A seed value is optional, but you should provide a value if you want reproducibility across runs of the same experiment.

Here we must replicate the findings.

Box 2: Mean Absolute Error

Scenario: Given a trained model and a test dataset, you must compute the Permutation Feature Importance scores of feature variables. You need to set up the Permutation Feature Importance module to select the correct metric to investigate the model’s accuracy and replicate the findings.

Regression. Choose one of the following: Precision, Recall, Mean Absolute Error , Root Mean Squared Error, Relative Absolute Error, Relative Squared Error, Coefficient of Determination

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/permutation-feature-importance

Box 1: Replace using MICE

Replace using MICE: For each missing value, this option assigns a new value, which is calculated by using a method described in the statistical literature as "Multivariate Imputation using Chained Equations" or "Multiple Imputation by Chained Equations". With a multiple imputation method, each variable with missing data is modeled conditionally using the other variables in the data before filling in the missing values.

Scenario: The AccessibilityToHighway column in both datasets contains missing values. The missing data must be replaced with new data so that it is modeled conditionally using the other variables in the data before filling in the missing values.

Box 2: Propagate

Cols with all missing values indicate if columns of all missing values should be preserved in the output.

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/clean-missing-data

In statistics, the Kendall rank correlation coefficient, commonly referred to as Kendall's tau coefficient (after the Greek letter τ), is a statistic used to measure the ordinal association between two measured quantities.

It is a supported method of the Azure Machine Learning Feature selection.

Scenario: When you train a Linear Regression module using a property dataset that shows data for property prices for a large city, you need to determine the best features to use in a model. You can choose standard metrics provided to measure performance before and after the feature importance process completes. You must ensure that the distribution of the features across multiple training models is consistent.

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/feature-selection-modules

Box 1: Mutual Information.

The mutual information score is particularly useful in feature selection because it maximizes the mutual information between the joint distribution and target variables in datasets with many dimensions.

Box 2: MedianValue

MedianValue is the feature column, , it is the predictor of the dataset.

Scenario: The MedianValue and AvgRoomsinHouse columns both hold data in numeric format. You need to select a feature selection algorithm to analyze the relationship between the two columns in more detail.

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/filter-based-feature-selection

Box 1: forcasting

Task: The type of task to run. Values can be 'classification', 'regression', or 'forecasting' depending on the type of automated ML problem to solve.

Box 2: temperature

The training data to be used within the experiment. It should contain both training features and a label column (optionally a sample weights column).

Box 3: observation_time

time_column_name: The name of the time column. This parameter is required when forecasting to specify the datetime column in the input data used for building the time series and inferring its frequency. This setting is being deprecated. Please use forecasting_parameters instead.

Box 4: 7

"predicts temperature over the next seven days"

max_horizon: The desired maximum forecast horizon in units of time-series frequency. The default value is 1.

Units are based on the time interval of your training data, e.g., monthly, weekly that the forecaster should predict out. When task type is forecasting, this parameter is required.

Box 5: 50

"For the initial round of training, you want to train a maximum of 50 different models."

Iterations: The total number of different algorithm and parameter combinations to test during an automated ML experiment.

Leave One Out (LOO) cross-validation

Setting K = n (the number of observations) yields n-fold and is called leave-one out cross-validation (LOO), a special case of the K-fold approach.

LOO CV is sometimes useful but typically doesn’t shake up the data enough. The estimates from each fold are highly correlated and hence their average can have high variance.

This is why the usual choice is K=5 or 10. It provides a good compromise for the bias-variance tradeoff.

Instead use only auth_enabled = TRUE

Note: Key-based authentication.

Web services deployed on AKS have key-based auth enabled by default. ACI-deployed services have key-based auth disabled by default, but you can enable it by setting auth_enabled = TRUE when creating the ACI web service. The following is an example of creating an ACI deployment configuration with key-based auth enabled.

deployment_config <- aci_webservice_deployment_config(cpu_cores = 1,

memory_gb = 1,

auth_enabled = TRUE)

Azure Machine Learning compute cluster is a managed-compute infrastructure that allows you to easily create a single or multi-node compute. The compute is created within your workspace region as a resource that can be shared with other users in your workspace. The compute scales up automatically when a job is submitted, and can be put in an Azure Virtual Network.

Box 1:Yes

ComputeTargetException class: An exception related to failures when creating, interacting with, or configuring a compute target. This exception is commonly raised for failures attaching a compute target, missing headers, and unsupported configuration values.

Create(workspace, name, provisioning_configuration)

Provision a Compute object by specifying a compute type and related configuration.

This method creates a new compute target rather than attaching an existing one.

Box 2: Yes

Box 3: No

The line before print('Step1') will fail.

The evaluation metrics available for binary classification models are: Accuracy, Precision, Recall, F1 Score, and AUC.

Note: A very natural question is: ‘Out of the individuals whom the model, how many were classified correctly (TP)?’

This question can be answered by looking at the Precision of the model, which is the proportion of positives that are classified correctly.

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio/evaluate-model-performance

How to work around or solve common Docker deployment errors with Azure Container Instances (ACI) and Azure Kubernetes Service (AKS) using Azure Machine Learning.

The recommended and the most up to date approach for model deployment is via the Model.deploy() API using an Environment object as an input parameter. In this case our service will create a base docker image for you during deployment stage and mount the required models all in one call. The basic deployment tasks are:

1. Register the model in the workspace model registry.

2. Define Inference Configuration:

a. Create an Environment object based on the dependencies you specify in the environment yaml file or use one of our procured environments.

b. Create an inference configuration (InferenceConfig object) based on the environment and the scoring script.

3. Deploy the model to Azure Container Instance (ACI) service or to Azure Kubernetes Service (AKS).

Python printing/logging example:

logging.info(message)

Destination: Driver logs, Azure Machine Learning designer

Box 1: PCA(n_components = 10)

Need to reduce the dimensionality of the feature set to 10 features in both training and testing sets.

Example:

from sklearn.decomposition import PCA

pca = PCA(n_components=2) ;2 dimensions

principalComponents = pca.fit_transform(x)

Box 2: pca

fit_transform(X[, y])fits the model with X and apply the dimensionality reduction on X.

Box 3: transform(x_test)

transform(X) applies dimensionality reduction to X.

References:

https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html

If you have an existing Conda environment on your local computer, then you can use the service to create an environment object. By using this strategy, you can reuse your local interactive environment on remote runs.

Box 1: No

Environment is a required parameter. The environment to use for the run. If no environment is specified, azureml.core.runconfig.DEFAULT_CPU_IMAGE will be used as the Docker image for the run.

The following example shows how to instantiate a new environment.

from azureml.core import Environment

myenv = Environment(name="myenv")

Box 2: Yes

Parameter compute_target: The compute target where training will happen. This can either be a ComputeTarget object, the name of an existing ComputeTarget, or the string "local". If no compute target is specified, your local machine will be used.

Box 3: Yes

Parameter source_directory. A local directory containing code files needed for a run.

Parameter script. The file path relative to the source_directory of the script to be run.

Box 1: 4

Choose the one which has lower training and validation error and also the closest match.

Minimize variance (difference between validation error and train error).

Box 2: 5

Minimize variance (difference between validation error and train error).

References:

https://machinelearningmastery.com/tactics-to-combat-imbalanced-classes-in-your-machine-learning-dataset/

Instead use Permutation Feature Importance Explainer (PFI).

Note 1:

Note 2: Permutation Feature Importance Explainer (PFI): Permutation Feature Importance is a technique used to explain classification and regression models. At a high level, the way it works is by randomly shuffling data one feature at a time for the entire dataset and calculating how much the performance metric of interest changes. The larger the change, the more important that feature is. PFI can explain the overall behavior of any underlying model but does not explain individual predictions.

Box 1: Split rows

Use the Split Rows option if you just want to divide the data into two parts. You can specify the percentage of data to put in each split, but by default, the data is divided 50-50.

You can also randomize the selection of rows in each group, and use stratified sampling. In stratified sampling, you must select a single column of data for which you want values to be apportioned equally among the two result datasets.

Box 2: 0.75

If you specify a number as a percentage, or if you use a string that contains the "%" character, the value is interpreted as a percentage. All percentage values must be within the range (0, 100), not including the values 0 and 100.

Box 3: Yes

To ensure splits are balanced.

Box 4: No

If you use the option for a stratified split, the output datasets can be further divided by subgroups, by selecting a strata column.

Scenario:

Experiments for local crowd sentiment models must combine local penalty detection data.

Crowd sentiment models must identify known sounds such as cheers and known catch phrases. Individual crowd sentiment models will detect similar sounds.

Note: Evaluate the changed in correlation between model error rate and centroid distance

In machine learning, a nearest centroid classifier or nearest prototype classifier is a classification model that assigns to observations the label of the class of training samples whose mean (centroid) is closest to the observation.

References:

https://en.wikipedia.org/wiki/Nearest_centroid_classifier

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/sweep-clustering

Split Data partitions the rows of a dataset into two distinct sets.

The Relative Expression Split option in the Split Data module of Azure Machine Learning Studio is helpful when you need to divide a dataset into training and testing datasets using a numerical expression.

Relative Expression Split: Use this option whenever you want to apply a condition to a number column. The number could be a date/time field, a column containing age or dollar amounts, or even a percentage. For example, you might want to divide your data set depending on the cost of the items, group people by age ranges, or separate data by a calendar date.

Scenario:

Local market segmentation models will be applied before determining a user’s propensity to respond to an advertisement.

The distribution of features across training and production data are not consistent

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/split-data

The linear discriminant analysis method works only on continuous variables, not categorical or ordinal variables.

Linear discriminant analysis is similar to analysis of variance (ANOVA) in that it works by comparing the means of the variables.

Scenario:

Data scientists must build notebooks in a local environment using automatic feature engineering and model building in machine learning pipelines.

Experiments for local crowd sentiment models must combine local penalty detection data.

All shared features for local models are continuous variables.

Step 1: Implement a K-Means Clustering model

Step 2: Use the cluster as a feature in a Decision jungle model.

Decision jungles are non-parametric models, which can represent non-linear decision boundaries.

Step 3: Use the raw score as a feature in a Score Matchbox Recommender model

The goal of creating a recommendation system is to recommend one or more "items" to "users" of the system. Examples of an item could be a movie, restaurant, book, or song. A user could be a person, group of persons, or other entity with item preferences.

Scenario:

Ad response rated declined.

Ad response models must be trained at the beginning of each event and applied during the sporting event.

Market segmentation models must optimize for similar ad response history.

Ad response models must support non-linear boundaries of features.

References:

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/multiclass-decision-jungle

https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/score-matchbox-recommender

Step 1: Define a cross-entropy function activation

When using a neural network to perform classification and prediction, it is usually better to use cross-entropy error than classification error, and somewhat better to use cross-entropy error than mean squared error to evaluate the quality of the neural network.

Step 2: Add cost functions for each target state.

Step 3: Evaluated the distance error metric.

References:

https://www.analyticsvidhya.com/blog/2018/04/fundamentals-deep-learning-regularization-techniques/

Scenario:

Performance curves of current and proposed cost factor scenarios are shown in the following diagram:

The ad propensity model uses a cut threshold is 0.45 and retrains occur if weighted Kappa deviated from 0.1 +/- 5%.

Post batch normalization statistics (PBN) is the Microsoft Cognitive Toolkit (CNTK) version of how to evaluate the population mean and variance of Batch Normalization which could be used in inference Original Paper.

In CNTK, custom networks are defined using the BrainScriptNetworkBuilder and described in the CNTK network description language "BrainScript."

Scenario:

Local penalty detection models must be written by using BrainScript.

References:

https://docs.microsoft.com/en-us/cognitive-toolkit/post-batch-normalization-statistics

Box 1: import pytorch as deeplearninglib

Box 2: ..DistributedSampler(Sampler)..

DistributedSampler(Sampler):

Sampler that restricts data loading to a subset of the dataset.

It is especially useful in conjunction with class:`torch.nn.parallel.DistributedDataParallel`. In such case, each process can pass a DistributedSampler instance as a DataLoader sampler, and load a subset of the original dataset that is exclusive to it.

Scenario: Sampling must guarantee mutual and collective exclusively between local and global segmentation models that share the same features.

Box 3: optimizer = deeplearninglib.train. GradientDescentOptimizer(learning_rate=0.10)