AutoML on the Cloud

By Hemanka Sarmah

• With the tremendous amount of data being generated. Businesses are changing the way they used to work.
• Companies are getting data-driven and revolutionizing the way they used to work by including ML and AI that automates most of the time-consuming tasks.

Also, Check out our Article on:

Introduction to AutoML-The future of industry ML execution
Applying AutoML (Part-1) using Auto-Sklearn
Applying AutoML(Part-2) with MLBox
Applying AutoML (Part-3) with TPOT
Applying AutoML (Part-4) using H2O
Automated Hyperparameter tuning

What is cloud computing and how is it affecting businesses?

Cloud Computing is referred to as the availability of computer resources/infrastructures over the internet for flexible, faster, and economical uses.

Businesses are shifting to the cloud because:

• Cloud gives the adaptability that assists clients with meeting their requirements.
• Cloud enables quick analysis and process of huge amounts of data, this can be done by multiple internal and external resources.
• It increases collaboration between team members as it enables people to edit files in realtime and access them remotely from anywhere in the world.
• Regular software and security updates.
• It helps save time and effort.

Benefits of using Cloud for ML

• Cloud provides Automated ML that has no code or low code features that enables people with less expertise to apply machine learning.
• Access to State-of-art models(Transfer Learning) that helps solve problems in no time.
• It makes it easy for businesses to experiment with ML as per the business problem.
• Scaling up as demand increases are easy.
• It leverages the speed and power of GPU for training models without any investment in hardware.

Google was the first to launch AutoML Vision on Cloud.

Invented by Quoc Le, behind google’s AutoML is its engine called Neural Architecture Search.
Quoc Le co-founded Google Brain along with Andrew NG and Jeff Dean.

AI Tools Available.

• Chatbot
• Image Classifier
• Speech Recognition
• and much more.

Cloud Giants like AWS, Microsoft Azure, and Google Cloud Platform offers many such tools.

Google Cloud Platform

GCP offers an AI platform where you can build and deploy models easily.

• The AI Platform is a fully managed, end-to-end platform for data science and machine learning.
• It offers code-based and no-code tools that can be put into production.
• AI Platform is targeted at audiences with every skill level.
• AI Platform makes it easy for developers, data scientists, and data engineers to streamline and scale their ML workflows.

The end-to-end ML life cycle in GCP includes:

• Data Preparation
• Model Building
• Model Validating
• Model Deployment

AI Tools Available in GCP

1. AI Explanations:
   Understand how each feature in your input data contributed to the model’s outputs.
2. AutoML:
   Easily develop high-quality custom machine learning models without writing training routines. Powered by Google’s state-of-the-art transfer learning and hyperparameter search technology.
3. Continuous Evaluation:
   Obtain metrics about the performance of your models in production. Compare predictions with ground truth labels to gain continual feedback and optimize model performance over time.
4. Data Labelling Service:
   Get highly accurate labels from human labelers for better machine learning models.
5. Deep Learning Containers:
   Quickly build and deploy models in a portable and consistent environment for all your AI applications.
6. Deep Learning VM Image:
   Instantiate a VM image containing the most popular AI frameworks on a Compute Engine instance without worrying about software compatibility.
7. Notebooks:
   Create, manage, and connect to VMs with JupyterLab, the standard data scientist workbench. VMs come with pre-installed deep learning frameworks and libraries.
8. Pipelines:
   Implement MLOps by orchestrating the steps in your ML workflow as a pipeline without the difficulty of setting up Kubeflow Pipelines with TensorFlow Extended (TFX).
9. Prediction:
   Easily deploy your models to manage, scalable endpoints for online or batch predictions.
10. TensorFlow Enterprise:
    Easily develop and deploy TensorFlow models on Google Cloud with enterprise-grade support and cloud-scale performance.
11. Training:
    Train any models in any framework on any hardware, from single machines to large clusters with multiple accelerators.
12. Vizier:
    Optimize your model’s output by intelligently tuning hyperparameters.
13. What-If Tool:
    Visualize your datasets and probe your model to better understand its behavior with an interactive visual interface.

Advantages

• Allows you to create customized deep learning models without knowing any data science
• Creates language translation models for niche areas
• Creates custom text classifiers
• Creates custom image classifiers
• Runs in the Google cloud

Disadvantages

• Does not run on-premises.

Amazon Web Services

• Amazon SageMaker is a fully managed service that provides every developer and data scientist with the ability to build, train, and deploy ML models at scale.
• It removes the complexity from each step of the ML workflow so you can more easily deploy your ML use cases, anything from predictive maintenance to computer vision to predicting customer behaviors.
• Hundreds of models and algorithms are available at the AWS marketplace.
• AWS offers Sagemaker and Sagemaker Autopilot for low code and no code usage of AutoML.

Services on AWS Sagemaker:

• Amazon Sagemaker Autopilot:
  → Amazon SageMaker Autopilot automatically builds, trains, and tunes the best machine learning models based on your data, while allowing you to maintain full control and visibility.
  → With SageMaker Autopilot, you simply provide a tabular dataset and select the target column to predict, which can be a number or a category.
• Amazon Sagemaker Ground Truth:
  → Amazon SageMaker Ground Truth is a fully managed data labeling service that makes it easy to build highly accurate training datasets for machine learning.
  → Get started with labeling your data in minutes through the SageMaker Ground Truth console using custom or built-in data labeling workflows.
• Amazon SageMaker JumpStart:
  → SageMaker JumpStart provides a set of solutions for the most common use cases that can be deployed readily with just a few clicks.
  → The solutions are fully customizable and showcase the use of AWS CloudFormation templates and reference architectures so you can accelerate your ML journey.
  → Amazon SageMaker JumpStart also supports one-click deployment and fine-tuning of more than 150 popular open-source models such as natural language processing, object detection, and image classification models.
• Amazon SageMaker Data Wrangler:
  → With SageMaker Data Wrangler, you can simplify the process of data preparation and feature engineering, and complete each step of the data preparation workflow, including data selection, cleaning, exploration, and visualization from a single visual interface.
  → SageMaker Data Wrangler contains over 300 built-in data transformations so you can quickly normalize, transform, and combine features without having to write any code.
  → Once your data is prepared, you can build fully automated ML workflows with Amazon SageMaker Pipelines and save them for reuse in the Amazon SageMaker Feature Store.
• Amazon SageMaker Clarify:
  → Provides machine learning developers with greater visibility into their training data and models so they can identify and limit bias and explain predictions.

Advantages

• Amazon’s own pre-built models have been highly optimized to run on AWS services.
• you do not need to install TensorFlow or other common libraries separately.
  AWS Sagemaker supports Tensorflow, PyTorch, and MXNet
• you can link your Github account with these notebooks — so that you can keep committing to the Github repo as you build the model — no need to worry about downloading and uploading files for version control

Disadvantages

• AWS sets default limits on resources which vary from region to region.
• AWS does have general cloud computing issues when you move to a cloud such as downtime, limited control, and backup protection.

Microsoft Azure

• Best-in-class support for open-source frameworks and languages including MLflow, Kubeflow, ONNX, PyTorch, TensorFlow, Python, and R.
• It has Azure Cognitive Services.
• Cognitive Services brings AI within reach of every developer without requiring machine-learning expertise.
• A Cognitive Service provides a trained model for you. This brings data and an algorithm together, available from a REST API(s) or SDK.
• You can implement this service within minutes, depending on your scenario.
• A Cognitive Service provides answers to general problems such as key phrases in text or item identification in images.
• All it takes is an API call to embed the ability to see, hear, speak, search, understand, and accelerate decision-making into your apps.
• To know about all the API's check out this Link

Advantages

• Get started quickly — no machine-learning expertise required
• Apply AI to more scenarios with the most comprehensive portfolio of domain-specific AI capabilities on the market.
• Microsoft’s AI capabilities in vision, speech, and language are equal to those of humans.
• The container-based architecture enables flexible deployment from the cloud that doesn’t require AI expertise.

Disadvantages

• Products are available only in specific Regions.
• A hybrid network and an integration strategy should be implemented before the start of development

Also, Check out our Article on:

Introduction to AutoML-The future of industry ML execution
Applying AutoML (Part-1) using Auto-Sklearn
Applying AutoML(Part-2) with MLBox
Applying AutoML (Part-3) with TPOT
Applying AutoML (Part-4) using H2O
Automated Hyperparameter tuning

Follow us for more upcoming future articles related to Data Science, Machine Learning, and Artificial Intelligence.

Also, Do give us a Clap👏 if you find this article useful as your encouragement catalyzes inspiration for and helps to create more cool stuff like this.

Visit us on https://www.insaid.co/