Practical AI for Cybersecurity

The world of cybersecurity and the landscape that it possesses is changing on a dynamic basis. It seems like that hardly one threat vector is launched, new variants of it are already on the way. IT Security teams in businesses and corporations are struggling daily to fight off any cyberthreats that they are experiencing. On top of this, they are also asked by their CIO or CISO to model what future Cyberattacks could potentially look like, and ways as to how the lines of defenses can be further enhanced. IT Security teams are overburdened and are struggling to find ways in order to keep up with what they are being asked to do. Trying to model the cyberthreat landscape is a very laborious process, because it takes a lot of time to analyze datasets from many intelligence feeds. What can be done to accomplish this Herculean task? The answer lies in Artificial Intelligence (AI). With AI, an IT Security team can model what the future Cyberthreat landscape could potentially look like in just a matter of minutes. As a result, this gives valuable time for them not only to fight off the threats that they are facing, but to also come up with solutions for the variants that will come out later. Practical AI for Cybersecurity explores the ways and methods as to how AI can be used in cybersecurity, with an emphasis upon its subcomponents of machine learning, computer vision, and neural networks. The book shows how AI can be used to help automate the routine and ordinary tasks that are encountered by both penetration testing and threat hunting teams. The result is that security professionals can spend more time finding and discovering unknown vulnerabilities and weaknesses that their systems are facing, as well as be able to come up with solid recommendations as to how the systems can be patched up quickly. -- Provided by publisher 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