AIPrimer.AI
  • 🚦AI Primer In Transportation
  • CHAPTER 1 - INTRODUCTION TO MACHINE LEARNING
    • Machine Learning in Transportation
    • What is Machine Learning?
    • Types of Machine Learning
      • Supervised Learning
      • Unsupervised Learning
      • Semi-supervised Learning
      • Reinforced Learning
    • Fundamental concepts of machine learning
      • Model Training and Testing
      • Evaluating the Model’s Prediction Accuracy
      • The Underfitting and Overfitting Problems
      • Bias-Variance Tradeoff in Overfitting
      • Model Validation Techniques
      • Hyperparameter Tuning
      • Model Regularization
      • The Curse of Ddimensionality
    • Machine Learning versus Statistics
  • CHAPTER 2 - SUPERVISED METHODS
    • Supervised Learning_Complete Draft
    • K-Nearest Neighbor (KNN) Algorithm
    • Tree-Based Methods
    • Boosting
    • Support Vector Machines (SVMs)
  • CHAPTER 3 - UNSUPERVISED LEARNING
    • Principal Component Analysis
      • How Does It Work?
      • Interpretation of PCA result
      • Applications in Transportation
    • CLUSTERING
      • K-MEANS
      • SPECTRAL CLUSTERING
      • Hierarchical Clustering
    • REFERENCE
  • CHAPTER 4 - NEURAL NETWORK
    • The Basic Paradigm: Multilayer Perceptron
    • Regression and Classification Problems with Neural Networks
    • Advanced Topologies
      • Modular Network
      • Coactive Neuro–Fuzzy Inference System
      • Recurrent Neural Networks
      • Jordan-Elman Network
      • Time-Lagged Feed-Forward Network
      • Deep Neural Networks
  • CHAPTER 5 - DEEP LEARNING
    • Convolutional Neural Networks
      • Introduction
      • Convolution Operation
      • Typical Layer Structure
      • Parameters and Hyperparameters
      • Summary of Key Features
      • Training of CNN
      • Transfer Learning
    • Recurrent Neural Networks
      • Introduction
      • Long Short-Term Memory Neural Network
      • Application in transportation
    • Recent Development
      • AlexNet, ZFNet, VggNet, and GoogLeNet
      • ResNet
      • U-Net: Full Convolutional Network
      • R-CNN, Fast R-CNN, and Faster R-CNN
      • Mask R-CNN
      • SSD and YOLO
      • RetinaNet
      • MobileNets
      • Deformable Convolution Networks
      • CenterNet
      • Exemplar Applications in Transportation
    • Reference
  • CHAPTER 6 - REINFORCEMENT LEARNING
    • Introduction
    • Reinforcement Learning Algorithms
    • Model-free v.s. Model-based Reinforcement Learning
    • Applications of Reinforcement Learning to Transportation and Traffic Engineering
    • REFERENCE
  • CHAPTER 7 - IMPLEMENTING ML AND COMPUTATIONAL REQUIREMENTS
    • Data Pipeline for Machine Learning
      • Introduction
      • Problem Definition
      • Data Ingestion
      • Data Preparation
      • Data Segregation
      • Model Training
      • Model Deployment
      • Performance Monitoring
    • Implementation Tools: The Machine Learning Ecosystem
      • Machine Learning Framework
      • Data Ingestion tools
      • Databases
      • Programming Languages
      • Visualization Tools
    • Cloud Computing
      • Types and Services
    • High-Performance Computing
      • Deployment on-premise vs on-cloud
      • Case Study: Data-driven approach for the implementation of Variable Speed Limit
      • Conclusion
  • CHAPTER 8 - RESOURCES
    • Mathematics and Statistics
    • Programming, languages, and software
    • Machine learning environments
    • Tools of the Trade
    • Online Learning Sites
    • Key Math Concepts
  • REFERENCES
  • IMPROVEMENT BACKLOG
Powered by GitBook
On this page
  1. CHAPTER 7 - IMPLEMENTING ML AND COMPUTATIONAL REQUIREMENTS
  2. Cloud Computing

Types and Services

PreviousCloud ComputingNextHigh-Performance Computing

Last updated 1 year ago

Cloud computing refers to the outsourcing of services such as software, analysis or database, etc., to a third-party organization popularly referred to as cloud service provider. The cloud service provider offers solutions for expanding services on a global scale with universal access to users via the internet. The service is hosted in one of the many data centers, owned by these organizations, in the world. Cloud computing is preferred for its workload scalability, flexibility in pay, accessibility and high performance in terms of processing power. Based on the user’s objective they can choose to set up their own cloud base or seek assistance from a cloud service provider. After choosing the right cloud base for their project they can choose the type of service they would like to build and offer.

There are three different cloud bases, a user can choose from for deployment:

  1. Public cloud – Public cloud base is completely owned and operated by a third-party organization such as Amazon Web Services, Microsoft Azure, etc. When a user’s service is deployed in public cloud, the entire infrastructure, hardware and software of the service is owned by the service provider and the user can access their application and data using a web browser.

  2. Private cloud – This is a cloud base hosted by a business organization for its own private use. This cloud can be physically located close to the organization's data center or may be deployed by a third-party organization on a private network. The complete resources with infrastructure are owned by the business organization.

  3. Hybrid cloud – If a combination of cloud bases (public and private) is used for deployment, it is called a hybrid cloud. This cloud base combines the best of both types allowing for sharing of data and applications between them. It provides greater flexibility, optimizes the existing infrastructure and security.

Cloud-based solutions can be broadly categorized into three categories based on the type of assistance rendered (Figure 5-13):

  1. Software as a service (SaaS) allows users to host and access software applications. The complete software from infrastructure to its data is hosted in a data center owned by the cloud service provider. To access the hosted software users can simply log in to their account with a viable internet connection. All the information collected by the software is managed securely by the service provider as per the terms of their service agreement. This is an effective technique for organizations to launch and use their software with a minimal upfront cost.

  2. Platform as a service (PaaS) provides the tools and resources required for software development. The service provider assists with libraries, codes, programming environments and back-end infrastructures such as servers and storage. These resources are required throughout the lifecycle of an application assisting in building, testing, deploying, managing and updating the application.

  3. Infrastructure as a service (IaaS) provides the computing power and storage which the user can access via the internet. Based on user preference, the service provider allows them storage capacity which they can access and configure as seen fit to their objective.

There is another special type of cloud computing called serverless computing where the entire infrastructure setup, provisioning, and management is handled by the service provider. The infrastructure is capable of dynamically scaling up or down as per the demand of workload. The major advantage of serverless computing is the reduction of workload and resources for the user to manage the infrastructure. As the service provider handles these areas, the user is left with time to focus and concentrate on their business logic.

Figure 5-13 Cloud Services

Pros and Cons

Advantages:

  • Cost: Cloud computing services offer a pay-as-you-use type of scheme which is highly flexible allowing the user to pay for the services they use. For example, in PaaS, users can pick and choose the tools they require and pay just for that. In the case of IaaS, users can pay per memory space allotted and later change their pricing schemes based on the increase or decrease in the demand. This flexibility with payments helps organization especially startups to invest their funds efficiently and run a cost-effective business.

  • Scalability of Workload: Cloud service providers such as Salesforce, Amazon, Google, and Microsoft have a large network of secure data centers deployed across the world. They have the capability and resources to increase the power, storage, and bandwidth as per the user’s needs. Hence, the user can increase or decrease their workload and tailor the cloud services as per the demand without any wastage of resources.

  • Storage: Cloud computing offers reliable storage and backup of data. Data stored in the cloud is available as many copies in different data centers across the globe. Hence, even if the local system crashes, data in the cloud is safely backed up and can be easily retrieved.

  • IT Support: There is abundant tech support available with the service provider which eliminates the need for expert knowledge or an in-house IT staff for repair and maintenance. This is a huge relief for small businesses and startups who might not have the funds to support a large staff group.

  • Security: Cloud service providers invest in setting up technologies and policies to secure their user’s data. They work hard at meeting the compliance standards for security which are monitored through independent third-party audits. Through these audits, the cloud service provider showcases his capabilities to ensure security compliance that meets stringent standards. They also set up strong firewall systems that can filter requests coming into the system and keep out the malicious content. To maintain a long sustaining business in cloud computing, service providers ensure expert security systems to protect data and win the costumer’s vote of confidence.

  • Universal access: The entire service is hosted in the cloud hence internet is the only requirement for all members of the team to obtain access to the components of the data. Service providers can scale elastically providing the services to its users from around the world. Along with improving user accessibility, it also allows for teams from different geographical regions to collaborate with ease.

Disadvantages:

  • Connectivity: The key component of cloud computing is a viable internet connection via which the user can access the cloud base. When the connection is weak, the user will be completely blocked out without access to their own data or application. In developing countries, this could pose a huge problem to business productivity.

  • Migration: Each cloud computing service is accompanied by its own set of rules and syntax. Sometimes, when the user needs to terminate the service and move data to a different platform it can pose problems. The user will have to go through the trouble of tailoring the entire application for the requirements of the new cloud service for its smooth functioning. Hence, the migration of data and application across different cloud service providers can be difficult.

  • Security: When a cloud computing system is working online it is vulnerable to a range of attacks which could result in loss of data. Although the service providers are well equipped to counter these attacks they haven’t been completely eradicated.