Practical AI for Cybersecurity

Cover
Half Title
Title Page
Copyright Page
Dedication
Table of contents
Acknowledgments
Notes on Contributors
Chapter 1 Artificial Intelligence
The Chronological Evolution of Cybersecurity
An Introduction to Artificial Intelligence
The Sub-Fields of Artificial Intelligence
Machine Learning
Neural Networks
Computer Vision
A Brief Overview of This Book
The History of Artificial Intelligence
The Origin Story
The Golden Age for Artificial Intelligence
The Evolution of Expert Systems
The Importance of Data in Artificial Intelligence
The Fundamentals of Data Basics
The Types of Data that are Available
Big Data
Understanding Preparation of Data
Other Relevant Data Concepts that are Important to Artificial Intelligence
Resources
Chapter 2 Machine Learning
The High Level Overview
The Machine Learning Process
Data Order
Picking the Algorithm
Training the Model
Model Evaluation
Fine Tune the Model
The Machine Learning Algorithm Classifications
The Machine Learning Algorithms
Key Statistical Concepts
The Deep Dive into the Theoretical Aspects of Machine Learning
Understanding Probability
The Bayesian Theorem
The Probability Distributions for Machine Learning
The Normal Distribution
Supervised Learning
The Decision Tree
The Problem of Overfitting the Decision Tree
The Random Forest
Bagging
The Naïve Bayes Method
The KNN Algorithm
Unsupervised Learning
Generative Models
Data Compression
Association
The Density Estimation
The Kernel Density Function
Latent Variables
Gaussian Mixture Models
The Perceptron
Training a Perceptron
The Boolean Functions
The Multiple Layer Perceptrons
The Multi-Layer Perceptron (MLP): A Statistical Approximator
The Backpropagation Algorithm
The Nonlinear Regression
The Statistical Class Descriptions in Machine Learning
Two Class Statistical Discrimination
Multiclass Distribution
Multilabel Discrimination
Overtraining
How a Machine Learning System can Train from Hidden, Statistical Representation
Autoencoders
The Word2vec Architecture
Application of Machine Learning to Endpoint Protection
Feature Selection and Feature Engineering for Detecting Malware
Common Vulnerabilities and Exposures (CVE)
Text Strings
Byte Sequences
Opcodes
API, System Calls, and DLLs
Entropy
Feature Selection Process for Malware Detection
Feature Selection Process for Malware Classification
Training Data
Tuning of Malware Classification Models Using a Receiver Operating Characteristic Curve
Detecting Malware after Detonation
Summary
Applications of Machine Learning Using Python
The Use of Python Programming in the Healthcare Sector
How Machine Learning is Used with a Chatbot
The Strategic Advantages of Machine Learning In Chatbots
An Overall Summary of Machine Learning and Chatbots
The Building of the Chatbot—A Diabetes Testing Portal
The Initialization Module
The Graphical User Interface (GUI) Module
The Splash Screen Module
The Patient Greeting Module
The Diabetes Corpus Module
The Chatbot Module
The Sentiment Analysis Module
The Building of the Chatbot—Predicting Stock Price Movements
The S&P 500 Price Acquisition Module
Loading Up the Data from the API
The Prediction of the Next Day Stock Price Based upon Today’s Closing Price Module
The Financial Data Optimization (Clean-Up) Module
The Plotting of SP500 Financial Data for the Previous Year + One Month
The Plotting of SP500 Financial Data for One Month
Calculating the Moving Average of an SP500 Stock
Calculating the Moving Average of an SP500 Stock for just a One Month Time Span
The Creation of the NextDayOpen Column for SP500 Financial Price Prediction
Checking for any Statistical Correlations that Exist in the NextDayOpen Column for SP500 Financial Price Prediction
The Creation of the Linear Regression Model to Predict Future SP500 Price Data
Sources
Application Sources
Chapter 3 The High Level Overview into Neural Networks
The High Level Overview into Neural Networks
The Neuron
The Fundamentals of the Artificial Neural Network (ANN)
The Theoretical Aspects of Neural Networks
The Adaline
The Training of the Adaline
The Steepest Descent Training
The Madaline
An Example of the Madaline: Character Recognition
The Backpropagation
Modified Backpropagation (BP) Algorithms
The Momentum Technique
The Smoothing Method
A Backpropagation Case Study: Character Recognition
A Backpropagation Case Study: Calculating the Monthly High and Low Temperatures
The Hopfield Networks
The Establishment, or the Setting of the Weights in the Hopfield Neural Network
Calculating the Level of Specific Network Stability in the Hopfield Neural Network
How the Hopfield Neural Network Can Be Implemented
The Continuous Hopfield Models
A Case Study Using the Hopfield Neural Network: Molecular Cell Detection
Counter Propagation
The Kohonen Self-Organizing Map Layer
The Grossberg Layer
How the Kohonen Input Layers are Preprocessed
How the Statistical Weights are Initialized in the Kohonen Layer
The Interpolative Mode Layer
The Training of the Grossberg Layers
The Combined Counter Propagation Network
A Counter Propagation Case Study: Character Recognition
The Adaptive Resonance Theory
The Comparison Layer
The Recognition Layer
The Gain and Reset Elements
The Establishment of the ART Neural Network
The Training of the ART Neural Network
The Network Operations of the ART Neural Network
The Properties of the ART Neural Network
Further Comments on Both ART 1 & ART 2 Neural Networks
An ART 1 Case Study: Making Use of Speech Recognition
The Cognitron and the Neocognitron
The Network Operations of the Excitory and Inhibitory Neurons
For the Inhibitory Neuron Inputs
The Initial Training of the Excitory Neurons
Lateral Inhibition
The Neocognitron
Recurrent Backpropagation Networks
Fully Recurrent Networks
Continuously Recurrent Backpropagation Networks
Deep Learning Neural Networks
The Two Types of Deep Learning Neural Networks
The LAMSTAR Neural Networks
The Structural Elements of LAMSTAR Neural Networks
The Mathematical Algorithms That Are Used for Establishing the Statistical Weights for the Inputs and the Links in the ...
An Overview of the Processor in LAMSTAR Neural Networks
The Training Iterations versus the Operational Iterations
The Issue of Missing Data in the LAMSTAR Neural Network
The Decision-Making Process of the LAMSTAR Neural Network
The Data Analysis Functionality in the LAMSTAR Neural Network
Deep Learning Neural Networks—The Autoencoder
The Applications of Neural Networks
The Major Cloud Providers for Neural Networks
The Neural Network Components of the Amazon Web Services & Microsoft Azure
The Amazon Web Services (AWS)
The Amazon SageMaker
From the Standpoint of Data Preparation
From the Standpoint of Algorithm Selection, Optimization, and Training
From the Standpoint of AI Mathematical Algorithm and Optimizing
From the Standpoint of Algorithm Deployment
From the Standpoint of Integration and Invocation
The Amazon Comprehend
Amazon Rekognition
Amazon Translate
Amazon Transcribe
Amazon Textract
Microsoft Azure
The Azure Machine Learning Studio Interactive Workspace
The Azure Machine Learning Service
The Azure Cognitive Services
The Google Cloud Platform
The Google Cloud AI Building Blocks
Building an Application That Can Create Various Income Classes
Building an Application That Can Predict Housing Prices
Building an Application That Can Predict Vehicle Traffic Patterns in Large Cities
Building an Application That Can Predict E-Commerce Buying Patterns
Building an Application That Can Recommend Top Movie Picks
Building a Sentiment Analyzer Application
Application of Neural Networks to Predictive Maintenance
Normal Behavior Model Using Autoencoders
Wind Turbine Example
Resources
Chapter 4 Typical Applications for Computer Vision
Typical Applications for Computer Vision
A Historical Review into Computer Vision
The Creation of Static and Dynamic Images in Computer Vision (Image Creation)
The Geometric Constructs—2-Dimensional Facets
The Geometric Constructs—3-Dimensional Facets
The Geometric Constructs—2-Dimensional Transformations
The Geometric Constructs—3-Dimensional Transformations
The Geometric Constructs—3-Dimensional Rotations
Ascertaining Which 3-Dimensional Technique Is the Most Optimized to Use for the ANN System
How to Implement 3-Dimensional Images onto a Geometric Plane
The 3-Dimensional Perspective Technique
The Mechanics of the Camera
Determining the Focal Length of the Camera
Determining the Mathematical Matrix of the Camera
Determining the Projective Depth of the Camera
How a 3-Dimensional Image Can Be Transformed between Two or More Cameras
How a 3-Dimensional Image Can Be Projected into an Object-Centered Format
How to Take into Account the Distortions in the Lens of the Camera
How to Create Photometric, 3-Dimensional Images
The Lighting Variable
The Effects of Light Reflectance and Shading
The Importance of Optics
The Effects of Chromatic Aberration
The Properties of Vignetting
The Properties of the Digital Camera
Shutter Speed
Sampling Pitch
Fill Factor
Size of the Central Processing Unit (CPU)
Analog Gain
Sensor Noise
The ADC Resolution
The Digital Post-Processing
The Sampling of the 2-Dimensional or 3-Dimensional Images
The Importance of Color in the 2-Dimensional or 3-Dimensional Image
The CIE, RGB, and XYZ Theorem
The Importance of the L*a*b Color Regime for 2-Dimensional and 3-Dimensional Images
The Importance of Color-Based Cameras in Computer Vision
The Use of the Color Filter Arrays
The Importance of Color Balance
The Role of Gamma in the RGB Color Regime
The Role of the Other Color Regimes in 2-Dimensional and 3-Dimensional Images
The Role of Compression in 2-Dimensional and 3-Dimensional Images
Image Processing Techniques
The Importance of the Point Operators
The Importance of Color Transformations
The Impacts of Image Matting
The Impacts of the Equalization of the Histogram
Making Use of the Local-Based Histogram Equalization
The Concepts of Linear Filtering
The Importance of Padding in the 2-Dimensional or 3-Dimensional Image
The Effects of Separable Filtering
What the Band Pass and Steerable Filters Are
The Importance of the Integral Image Filters
A Breakdown of the Recursive Filtering Technique
The Remaining Operating Techniques That Can Be Used by the ANN System
An Overview of the Median Filtering Technique
A Review of the Bilateral Filtering Technique
The Iterated Adaptive Smoothing/Anisotropic Diffusion Filtering Technique
The Importance of the Morphology Technique
The Impacts of the Distance Transformation Technique
The Effects of the Connected Components
The Fourier Transformation Techniques
The Importance of the Fourier Transformation-Based Pairs
The Importance of the 2-Dimensional Fourier Transformations
The Impacts of the Weiner Filtering Technique
The Functionalities of the Discrete Cosine Transform
The Concepts of Pyramids
The Importance of Interpolation
The Importance of Decimation
The Importance of Multi-Level Representations
The Essentials of Wavelets
The Importance of Geometric-Based Transformations
The Impacts of Parametric Transformations
Resources
Chapter 5 Conclusion
Index