Real-Time Cloud Computing and Machine Learning Applications

Contents
Preface
Chapter 1
Introduction
	1.1. Overview of Cloud Computing, Its Benefits  and Applications
		1.1.1. Benefits of Cloud Computing
		1.1.2. Applications of Cloud Computing
			1.1.2.1. Cloud Application Development
			1.1.2.2. Cloud as an Enabler for Industry 4.0
			1.1.2.3. Cloud Radio Access Networks (C-RAN)
			1.1.2.4. Big Data Analytics
				Cloud Private Branch Exchange (PBX)
	1.2. Overview of Machine Learning and AI
		1.2.1. Benefits of Machine Learning and AI
		1.2.2. Applications of Machine Learning and AI
	1.3. Combining AI with Cloud Computing for  Real-Time Applications
	1.4. Book Overview
	References
Chapter 2
Cloud Computing Fundamentals
	2.1. Definitions of Cloud Computing
	2.2. Computing Paradigms
	2.3. Cloud Computing Architecture and  Enabling Technologies
	2.4. Cloud Computing Deployment Models and  Service Classes
		2.4.1. Deployment Models
		2.4.2. Service Classes
	2.5. Introduction to the Firebase Cloud Platform
		2.5.1. Firebase Cloud Database Configurations
		2.5.2. Creating a Firebase Project and Real-time Database
		2.5.3. Sending and Reading Data from the Database with an Android Application
	2.6. Application Hosting on Firebase Using Node.js
		2.6.1. Node.js Overview and Installation
		2.6.2. Hosting a Web Application Using Node.js on Firebase for Reading and Writing Data to the Firebase Real-Time Database
	2.7. Introduction to IBM Cloud Platform
		2.7.1. IBM Cloud Database Configurations
		2.7.2. Desktop Application to Send and Receive Data to the IBM Cloudant Database
		2.7.3. Mobile Application to Send and Receive Data to the IBM Cloudant Database
	2.8. Application Hosting on IBM Bluemix via the Eclipse IDE
		2.8.1. Creating a Desktop Application to Send a Request to the CalculateArea servlet
		2.8.2. Creating a Mobile Application to Send a Request to the CalculateArea Servlet
	References
Chapter 3
Machine Learning Algorithms
	3.1. Definition of AI, Machine Learning and Deep Learning
		3.1.1. Artificial Narrow Intelligence (ANI)
		3.1.2. Artificial General Intelligence (AGI)
		3.1.3. Artificial Super Intelligence (ASI)
	3.2. Overview of Machine Learning Algorithms
	3.3. Unsupervised Learning Algorithms
		3.3.1. Unsupervised Shallow Learning Models
			3.3.1.1. K-Means Clustering
			3.3.1.2. Hierarchical Clustering
			3.3.1.3. Gaussian Mixture Models
		3.3.2. Unsupervised Deep Learning Models
			3.3.2.1. Restricted Boltzmann Machine (RBM)
	3.4. Supervised Learning Algorithms
		3.4.1. Supervised Shallow Learning Models
			3.4.1.1. Simple Linear Regression
			3.4.1.2. Multiple Linear Regression
			3.4.1.3. Polynomial Regression
			3.4.1.4. Naïve Bayes
			3.4.1.5. K-Nearest Neighbour
		3.4.2. Supervised Deep Learning Models
			3.4.2.1. Multi-Layered Perceptrons
			3.4.2.2. Convolutional Neural Network
	3.5. Reinforcement Learning Algorithms
		3.5.1. Q-Learning
		3.5.2. SARSA
	3.6. Ensemble Learning Algorithms
		3.6.1. Random Forest
	3.7. Deploying Javascript Machine Learning Algorithms  on Firebase
		3.7.1. Main Layout of the Application
		3.7.2. Incorporating the KNN Flower Classification Link
		3.7.3. Incorporating the Regression Algorithm Link
		3.7.4. Incorporating the Clustering Algorithm Link
	References
Chapter 4
Data Capture and Client Architecture  for a Cloud-Based Real-Time Network Analytics System
	4.1. Overview of Machine Learning Algorithms for  Network Analytics
		4.1.1. Classification of Network Data
		4.1.2. Regression Analysis for Network Data
	4.2. Complete System Model of the Network  Analytics System
	4.3. Mobile Application for Network Data Capture  and Analytics
		4.3.1. Creating the Android project
		4.3.2. Adding Libraries
		4.3.3. Android Application Layout
			4.3.3.1. Visual Outlook on Final Application
			4.3.3.2. Building the Visuals of activity_main.xml
			4.3.3.3. Building the Visuals of Prediction.xml
			4.3.3.4. Building the Visuals for csvdownload.xml
		4.3.4. Declaring Global Variables in Main Activity
		4.3.5. Retrieving the Last Index from the Cloud
		4.3.6. Traffic Monitoring in the onCreate() Method
			4.3.6.1. Initialising Components
			4.3.6.2. Getting Initial Readings
			4.3.6.3. Monitor Button
			4.3.6.4. Stop Button
			4.3.6.5. Go to Analysis Button
			4.3.6.6. Go to Download Button
		4.3.7. Getting Network Parameters
			4.3.7.1. Getting Speed Data
			4.3.7.2. Getting Packet Data
			4.3.7.3. Getting Wi-Fi Data
		4.3.8. Building the Main Thread
			4.3.8.1. Creating a Thread with Runnable Class
			4.3.8.2. Fetching Network Data
			4.3.8.3. Updating UI and Live Monitor
			4.3.8.4. Pushing Values to the Local Server
			4.3.8.5. Pushing Values Directly to the Cloud
		4.3.9. Live Graph Plotting
			4.3.9.1. Initialising the Chart
			4.3.9.2. Creating and Populating the Spinner
			4.3.9.3. Applying the Adapter to the Spinner
			4.3.9.4. Getting Spinner Value as a String
			4.3.9.5. Programmatically Adding Data onto the Live Graph
		4.3.10. Performing Analytics Using Cloud Servlet or Local Server
			4.3.10.1. The “Predict” Button
			4.3.10.2. The “Classify” Button
		4.3.11. Downloading to .csv Files
			4.3.11.1. Declaring Global Variables
			4.3.11.2. Initialising Components in onCreate()
			4.3.11.3. Browse Button
			4.3.11.4. Download Last N Samples Button
			4.3.11.5. Download by a Specific Date
			4.3.11.6. Issuing the Directory Picker
			4.3.11.7. Verifying Read and Write Permissions
			4.3.11.8. Handling Permissions Request
			4.3.11.9. Fetching Values from Cloudant Database
		4.3.12. Setting Permissions in Manifest
		4.3.13. Testing the Mobile application
			4.3.13.1. Live Monitor
			4.3.13.2. Downloading Functionalities
			4.3.13.3. Performing Analytics
	4.4. Desktop Application for Network Data  Capture and Analytics
		4.4.1. Creating the NetBeans project
		4.4.2. Desktop Application Layout
			4.4.2.1. Visual Outlook on Final Application
			4.4.2.2. Adding a JFrame Form to the Project
			4.4.2.3. Adding Components to JFrame
		4.4.3. Renaming Components
		4.4.4. Adding Libraries
		4.4.5. Declaring Global Variables in Netmonitor.Java
		4.4.6. Creating Live Monitor Layout
		4.4.7. Retrieving the Last Index from the Cloud
		4.4.8. Start Button
			4.4.8.1. Retrieving the Number of Items
			4.4.8.2. Finding Network Interface in Use
			4.4.8.3. Getting Initial Parameters
			4.4.8.4. Updating GUI and Live Monitor
			4.8.4.5. Pushing Values to the Local Server
			4.8.4.6. Pushing Values to the Cloud
		4.4.9. Stop Button
		4.4.10. Clearing Graph
		4.4.11. Performing Analytics Using Servlet or Local Server
			4.4.11.1. Predict Button
			4.4.11.2. Classify Button
		4.4.12. Downloading to .csv Files
			4.4.12.1. Browse Button
			4.4.12.2. Download Last N Samples Button
			4.4.12.3. Download by Specific Date Button
			4.4.12.4. Fetching Values from Cloudant Database
		4.4.13. Testing the Desktop Application
			4.4.13.1. Live Monitor
			4.4.13.2. Downloading Functionalities
			4.4.13.3. Performing Analytics
	4.5. Cloud Database Configurations for Network Analytics
		4.5.1. Creating a Cloudant Database to Store Network Data
		4.5.2. Adding an Index to the Database
	References
Chapter 5
Server and Servlet Architectures  for a Cloud-Based Real-Time Network  Analytics System
	5.1. Local Server Implementation for Network  Data Capture and Forecasting
		5.1.1. Creating the NetBeans Project
		5.1.2. Local Server Layout
			5.1.2.1. Visual Outlook on Final Application
			5.1.2.2. Adding Components to JFrame
		5.1.3. Renaming Components
		5.1.4. Adding Libraries
		5.1.5. Declaring Global Variables in LocServer.java
		5.1.6. Creating Live Monitor Layout
		5.1.7. Retrieving the Last Index from the Cloud
		5.1.8. Local Monitoring Server Implementation
			5.1.8.1. Calling getNumberOfItems
			5.1.8.2. Server Thread for Android Values
			5.1.8.3. Server Thread for PC Values
			5.1.8.4. Stopping the Server
			5.1.8.5. Uploading Android Values to the Cloud
		5.1.9. Filling Localhost Databases
		5.1.10. Analytics
		5.1.11. Classification
			5.1.11.1. Filling Cloudant Databases
			5.1.11.2. Retrieving Pre-Labelled Values from Cloudant Database or Localhost
			5.1.11.3. Classifying Network Parameters Using K-Nearest Neighbour (KNN)
			5.1.11.4. Performing K-Nearest Neighbour (KNN)
			5.1.11.5. Classifying Network Parameters Using Multilayer Perceptron (MLP)
			5.1.11.6. Performing Multilayer Perceptron (MLP) Classification
		5.1.12. Regression
			5.1.12.1. Using the Sliding Window Method
			5.1.12.2. Filling Cloudant Databases
			5.1.12.3. Retrieving Streaming Values from Cloudant Database
			5.1.12.4. Multiple Linear Regression Models for Network data
			5.1.12.5. Performing Multiple Linear Regression (MLR)
			5.1.12.6. Multilayer Perceptron Models for Network data
			5.1.12.7. Performing Multilayer Perceptron (MLP) Regression
		5.1.13. Downloading to .csv Files
			5.1.13.1. Browse Button
			5.1.13.2. Download the Last N Samples Button
			5.1.13.2. Download by Specific Date Button
		5.1.14. Local Download and Analytics
		5.1.15. Making the GUI User-Friendly
	5.2. Testing the Local Server
		5.2.1. Local Live Monitoring
		5.2.2. Downloading Functionalities
		5.2.3. Performing Analytics
	5.3. Servlet Program for Network Analytics,  Monitoring and Data Retrieval
		5.3.1. Creating a Dynamic Web Project in Eclipse IDE
		5.3.2. Adding Libraries to the Project
		5.3.3. Creating a Servlet Class for Analytics
		5.3.4. Building the Analytics Servlet
			5.3.4.1. Getting Network Parameters from the URL
			5.3.4.2. Adding Analytics Methods in the Servlet Class
		5.3.5. Building the Live Graph Servlets
			5.3.5.1. livegraph.java servlet
			5.3.5.2. livegraphAndroid.java servlet
		5.3.6. Building the Downloading Servlets
			5.3.6.1. downloadValues.java
			5.3.6.2. downloadValuesAndroid.java
			5.3.6.3. downloaderDate.java
			5.3.6.4. downloaderDateAndroid.java
		5.3.7. Web Client Application for Network Monitoring and Analytics
			5.3.7.1. Creating index1.html and index1.js
			5.3.7.2. Adding HTML Codes for index1.html
			5.3.7.3. Adding JavaScript Methods for index1.js
			5.3.7.4. Creating and Populating index2.html and index2.js
		5.3.8. Hosting the Java Servlet on IBM Bluemix
		5.3.9. Testing the Web Client Applications
			5.3.9.1. Live Monitor
			5.3.9.2. Performing Analytics
			6.3.9.3. Downloading Functionalities
	References
Chapter 6
Data Capture and Client Architecture for a Cloud-Based Real-Time Weather Forecasting System
	6.1. Overview of Machine Learning Algorithms for  Weather Forecasting
		6.1.1. Introduction
		6.1.2. Related Works
	6.2. Complete System Model of the Weather  Forecasting System
		6.2.1. MySQL Database
		6.2.2. Local Server
		6.2.3. Mobile and Desktop Application
		6.2.4. Cloudant Database
		6.2.5. Cloud Hosted Servlet
		6.2.6. Locally Hosted Servlet
		6.2.7. Web Application
	6.3. Mobile Application for Weather Data Capture  and Analytics
		6.3.1. Creating a New Android Application in Android Studio
		6.3.2. Libraries Used in the Android Application
		6.3.3. Program Structure
		6.3.4. Building the Main Activity Layout
		6.3.5. MainActivity.java
		6.3.6. Adding the Imports
		6.3.7. Global Variables
		6.3.8. The onCreate () Method
		6.3.9. The Createset () Method
		6.3.10. The Addentry () Method
		6.3.11. Clearing Graph on Item Selected
		6.3.12. Querying Weather API & Sending Data to Local Server or Cloudant Database
		6.3.13. QueryAPI Class
		6.3.14. The OpenWeather API
		6.3.15. The doInBackground () method
		6.3.16. onPostExecute ()
		6.3.17. Sending the Data to a Local Server or Cloud Database
		6.3.18. Analysis Activity
		6.3.19. Analysis Class
		6.3.20. QueryCloudServlet Class
		6.3.21. QueryLocalServer Class
		6.3.22. Querying Cloudant for Weather Readings
		6.3.23. DownloadData Activity
		6.3.24. DownloadData class
			6.3.24.1. onCreate () Method
			6.3.24.2. convertToCSV () Method
			6.3.24.3. writeFileToStorage() Method
		6.3.25. QueryDateCloudAsync Task
		6.3.26. QueryDateLocalTask
		6.3.27. DataRetriever Class
			6.3.27.1. DataRetriever () Constructor
			6.3.27.2. getNoOfRows () Method
			6.3.27.3. getDatabaseArray () method
		6.3.28. Testing the Mobile Application
			6.3.28.1. Live Monitoring
			6.3.28.2. Downloading Weather Data
			6.3.28.3. Performing Analytics
	6.4. Desktop Application for Weather Data Capture  and Analytics
		6.4.1. Creating the Desktop Application in NetBeans
		6.4.2. Adding Libraries
		6.4.3. Program Structure
		6.4.4. The MainFrame Class
			6.4.4.1. Adding the Imports
			6.4.4.2. Global Variables
			6.4.4.3. The MainFrame Constructor
			6.4.4.4. Timer & TimerTask Class
			6.4.4.5. The Monitoring Process
			6.4.4.6. The sendData () Method
			6.4.4.7. The requestAnalyticsCloud () Method
			6.4.4.8. The requestAnalyticsLocal () Method
			6.4.4.9. The requestDataCloud () Method
			6.4.4.10. The requestDataLocal () Method
			6.4.4.11. The download_dateActionPerformed () Method
			6.4.4.12. The downloadSamplesActionPerformed () Method
			6.4.4.13. The w_varActionPerformed () Method
		6.4.5. Testing the Desktop Application
			6.4.5.1. Live Monitoring
			6.4.5.2. Downloading Weather Data
			6.4.5.3. Performing Analytics
	6.5. Cloud Database Configurations and Hosting for Weather Forecasting
		6.5.1. Creating a New Cloudant Database to Store Weather Observations
		6.5.2. Adding a New Index to the Database
	References
Chapter 7
Server and Servlet Architectures  for a Cloud-Based Real-Time Weather  Forecasting System
	7.1. Local Server Implementation for Weather  Data Capture and Forecasting
		7.1.1. Local Server Program Structure
		7.1.2. Libraries Used
		7.1.3. Creating a New Java Application with a GUI in NetBeans
		7.1.4. The MainFrame Class
			7.1.4.1. Adding the Imports
			7.1.4.2. Global Variables
			7.1.4.3. The MainFrame () Constructor
			7.1.4.4. The w_varActionPerformed() Method
			7.1.4.5. The sendData () Method
			7.1.4.6. The local_serverActionPerformed () Method
		7.1.5. SQLConnector Class
			7.1.5.1. SQLConnector () Constructor
			7.1.5.2. The fetchLastN () Method
			7.1.5.3. The fetchForDate () Method
			7.1.5.4. The insertIntoDatabase () Method
		7.1.6. DataRetriever Class
		7.1.7. Regression Class
		7.1.8. Forecasting with the Regression Class
			7.1.8.1. Multiple Linear Regression Models for Weather Data
			7.1.8.2. Polynomial Regression Models for Weather Data
		7.1.9. Regression Example using MLR & Polynomial Regression
			7.1.9.1. Regression using MLR Model
			7.1.9.2. Regression using Polynomial model
		7.1.10. Details on OLSMultipleLinearRegression and SimpleRegression class
		7.1.11. RegressionPredictor Class
		7.1.12. Training Neural Network for Regression
			7.1.12.1. Normalization of the Data
			7.1.12.2. Reshape The Data into the Appropriate Form
			7.1.12.3. MLP Model Creation, Training & Prediction
			7.1.12.4. CNN Model Creation, Training & Prediction
		7.1.13. The Analytics () Method
		7.1.14. The predictActionPerformed () Method
		7.1.15. Rounding the Output to One Decimal Place
		7.1.16. Converting a Double Array to Float
		7.1.17. Formatting the JSONObject Returned by Analytics into a Readable Form
		7.1.18. Saving Data from the Cloud in A CSV File to the Local Storage
	7.2. Testing the Local Server
		7.2.1. Live Monitoring
		7.2.2. Downloading Weather Data
		7.2.3. Performing Analytics
	7.3. Servlet Program for Weather Forecasting  and Data Retrieval
		7.3.1. Installing the IBM WebSphere Plugin
		7.3.2. Servlet Program Structure
		7.3.3. Creating a Dynamic Web Project in Eclipse
		7.3.4. Adding a Servlet to the Project
		7.3.5. The “Forecaster” Servlet
		7.3.6. The “Downloader” Servlet
		7.3.7. The “graphUpdater” Servlet
	7.4. Web Client Application for Weather Forecasting  and Analytics
		7.4.1. HTML Page for User Interaction
			7.4.1.1. The formRequestURL Method
			7.4.1.2. Updating Graph
			7.4.1.3. Performing Prediction
			7.4.1.4. Download Weather Data
		7.4.2. Hosting the Java Servlet Program on IBM Cloud
	7.5. Testing the Web Client
		7.5.1. Live Monitoring
		7.5.2. Downloading Weather Data
		7.5.3. Performing Analytics
	References
About the Authors
Index
Blank Page
Blank Page