Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

How Google Does Machine Learning > All > Flashcards

All Flashcards

1
Q

Where are CNNs most used?

A
  • image classification
  • object detection
  • recommender systems.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Where are RNNs most used?

A
  • sequence modeling,
  • next word prediction
  • translating sounds to words
  • human language translation
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Where are Sorting and clustering architectures most used?

A
  • anomaly detection
  • pattern recognition
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Where are GANs most used?

A
  • anomaly detection
  • pattern recognition
  • cybersecurity
  • self-driving cars
  • reinforced learning.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How Vertex AI can be used?

A

Vertex AI be can use to manage the following stages in the ML workflow:
- Create a data set and upload data.
- Train an ML model on your data,
- evaluate model accuracy
- tune hyperparameters and custom training only.
- Upload and store your model in Vertex AI
- Deploy your trained model to an endpoint for serving predictions.
- Send prediction requests to your endpoint,
- specify prediction traffic split in your endpoint,
- manage your models and endpoints.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

When is it better to choose AutoML?

A
  • create and train a model with minimal technical effort
  • to quickly prototype models
  • explore new datasets before investing in development
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

When is it better to choose Custom training?

A
  • need to create a training application optimized for your targeted outcome.
  • want to have complete control over training application functionality.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is Vertex AI Feature Store?

A

Vertex AI Feature Store is a fully managed repository where you can ingest, serve, and share ML feature values within your organization. It manages all of the underlying infrastructure for you.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is Vertex Labeling tasks are for?

A

Data labeling tasks let you request human labeling for a dataset that you plan to use to train the custom machine learning model.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is Vertex AI workbench?

A

Vertex AI workbench is a Jupyter notebook-based development environment for the entire data science workflow.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is Vertex AI Labeling tasks for?

A

Data labeling tasks let you request human labeling for a dataset that you plan to use to train the custom machine learning model.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What Vertex AI Workbench lets you do?

A

Vertex AI Workbench lets you
- access data,
- process data in a Dataproc cluster,
- train a model,
- share your results,
- and more.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is Vertex AI Pipelines for?

A

Vertex AI Pipelines helps you to automate, monitor, and govern your ML systems by orchestrating your ML workflow in a serverless manner and storing your workflow’s artifacts using Vertex ML metadata.

It allows you to automate, monitor, and experiment with interdependent parts of an ML workflow. ML pipelines are portable, scalable, and based on containers and each individual part of your pipeline workflow.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What options are available for Jupyter notebooks in Vertex AI Workbench?

A
  • managed notebooks
  • user-managed notebooks
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What are Managed notebooks?

A

Managed notebooks instances are Google-managed environments with integrations and features that help you set up and work in an end-to-end notebook-based production environment.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What are User-managed notebooks?

A

User-managed notebooks are Deep Learning VM Images instances that are heavily customizable and are ideal if you need a lot of control over your environment.

17
Q

What are the best practices for storing data on GCP?

A
18
Q

Why do you need to avoid storing data in block storage like network file systems or in virtual machine hard disks?

A

Those tools are harder to manage than Cloud Storage or BigQuery and often present challenges in shooting performance.

19
Q

When should you create a new feature instead of using Vertex AI Feature Store?

A
  • If Vertex AI Feature Store contains features that you want to use, fetch those features for your training labels using Vertex AI Feature Store’s batch serving capability.
  • Create a new feature. If Vertex AI Feature Store doesn’t have the features you need, create a new feature using data from your data lake. Fetch raw data from your data lake and write your scripts to perform the necessary feature processing and engineering.
20
Q

Why should you add your new feature to Vertex AI Feature Store?

A

By adding your new feature to Vertex AI Feature Store, you automatically have a solution to an online serving of the features for online prediction use cases, and you can share your features with others in the organization that may get value from it for their own ML models.

21
Q

When is it recommended to use Vertex AI model training?

A
  • use the training service for larger datasets or for distributed training.
  • to productionized training, even on small datasets if the training is carried out on a schedule or in response to the arrival of additional data
22
Q

What is Vertex AI TensorBoard?

A

Vertex AI TensorBoard is an enterprise-ready managed service.

It provides a cost-effective, secure solution that lets data scientists and ML researchers collaborate easily by making it seamless to track, compare and share their experiments.

Vertex AI TensorBoard lets you track experiment metrics, such as loss and accuracy, over time, visualize a model graph, project embeddings to a lower dimensional space, and much more.

23
Q

What is best practicein processing data?

A

As a best practice, use BigQuery to process tabular data,
and use Dataflow to process unstructured data.

24
Q

What are the advantages of using managed datasets??

A

Managed datasets
- enable you to create a clear link between your data and custom-trained models
- provide descriptive statistics
- provide automatic or manual splitting into train, test, and validation sets.

25
Q

When may you choose not to use managed datasets?

A
  • if you want more control over splitting your data in your training code
  • if lineage between your data and model isn’t critical to your application.
26
Q

What can be used if you need to perform transformations that are not expressible in Cloud SQL?

A

If you need to perform transformations that are not expressible in Cloud SQL, or are for streaming, you can use a combination of Dataflow and the pandas library.

27
Q

What can be used to check for biases in the models?

A

What-If Tool. It’s available for free, and you will be able to access from within TensorBoard.

28
Q

How to visualize a large dataset for inclusiveness?

A

Use an open-source data visualization tool - Facets.
It can help you make machine learning more inclusive.

How Google Does Machine Learning (1 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions