book

Machine and Deep Learning Using MATLAB

Name: Machine and Deep Learning Using MATLAB
Author: Kamal I. M. Al-Malah
ISBN: 9781394209088

by Kamal I. M. Al-Malah

October 2023

Intermediate to advanced

592 pages

16h 6m

English

Wiley

Read now

Unlock full access

Cover
Title Page
Copyright Page
Preface
About the Companion Website
1 Unsupervised Machine Learning (ML) Techniques
IntroductionSelection of the Right Algorithm in MLClassical Multidimensional Scaling of Predictors DataPrincipal Component Analysis (PCA)k-Means ClusteringDistance Metrics: Locations of Cluster CentroidsReplicationsGaussian Mixture Model (GMM) ClusteringOptimum Number of GMM ClustersObservations and Clusters VisualizationEvaluating Cluster QualitySilhouette PlotsHierarchical ClusteringStep 1 – Determine Hierarchical StructureStep 2 – Divide Hierarchical Tree into ClustersPCA and Clustering: Wine QualityFeature Selection Using Laplacian (fsulaplacian) for Unsupervised LearningCHW 1.1 The Iris Flower Features DataCHW 1.2 The Ionosphere Data FeaturesCHW 1.3 The Small Car DataCHW 1.4 Seeds Features Data
2 ML Supervised Learning: Classification Models
Fitting Data Using Different Classification ModelsCustomizing a ModelCreating Training and Test DatasetsPredicting the ResponseEvaluating the Classification ModelKNN Model for All Categorical or All Numeric Data TypeKNN Model: Heart Disease Numeric DataViewing the Fitting Model PropertiesThe Fitting Model: Number of Neighbors and Weighting FactorThe Cost Penalty of the Fitting ModelKNN Model: Red Wine DataUsing MATLAB Classification LearnerBinary Decision Tree Model for Multiclass Classification of All Data TypesClassification Tree Model: Heart Disease Numeric Data TypesClassification Tree Model: Heart Disease All Predictor Data TypesNaïve Bayes Classification Model for All Data TypesFitting Heart Disease Numeric Data to Naïve Bayes ModelFitting Heart Disease All Data Types to Naïve Bayes ModelDiscriminant Analysis (DA) Classifier for Numeric Predictors OnlyDiscriminant Analysis (DA): Heart Disease Numeric PredictorsSupport Vector Machine (SVM) Classification Model for All Data TypesProperties of SVM ModelSVM Classification Model: Heart Disease Numeric Data TypesSVM Classification Model: Heart Disease All Data TypesMulticlass Support Vector Machine (fitcecoc) ModelMulticlass Support Vector Machines Model: Red Wine DataBinary Linear Classifier (fitclinear) to High-Dimensional DataCHW 2.1 Mushroom Edibility DataCHW 2.2 1994 Adult Census Income DataCHW 2.3 White Wine ClassificationCHW 2.4 Cardiac Arrhythmia DataCHW 2.5 Breast Cancer Diagnosis
3 Methods of Improving ML Predictive Models
Accuracy and Robustness of Predictive ModelsEvaluating a Model: Cross-ValidationCross-Validation Tune-up ParametersPartition with K-Fold: Heart Disease Data ClassificationReducing Predictors: Feature Transformation and SelectionFactor AnalysisFeature Transformation and Factor Analysis: Heart Disease DataFeature SelectionFeature Selection Using predictorImportance Function: Health Disease DataSequential Feature Selection (SFS): sequentialfs Function with Model Error HandlerAccommodating Categorical Data: Creating Dummy VariablesFeature Selection with Categorical Heart Disease DataEnsemble LearningCreating Ensembles: Heart Disease DataEnsemble Learning: Wine Quality ClassificationImproving fitcensemble Predictive Model: Abalone Age PredictionImproving fitctree Predictive Model with Feature Selection (FS): Credit Ratings DataImproving fitctree Predictive Model with Feature Transformation (FT): Credit Ratings DataUsing MATLAB Regression LearnerFeature Selection and Feature Transformation Using Regression Learner AppFeature Selection Using Neighborhood Component Analysis (NCA) for Regression: Big Car DataCHW 3.1 The Ionosphere DataCHW 3.2 Sonar DatasetCHW 3.3 White Wine ClassificationCHW 3.4 Small Car Data (Regression Case)
4 Methods of ML Linear Regression
IntroductionLinear Regression ModelsFitting Linear Regression Models Using fitlm FunctionHow to Organize the Data?Results Visualization: Big Car DataFitting Linear Regression Models Using fitglm FunctionNonparametric Regression Modelsfitrtree Nonparametric Regression Model: Big Car DataSupport Vector Machine, fitrsvm, Nonparametric Regression Model: Big Car DataNonparametric Regression Model: Gaussian Process Regression (GPR)Regularized Parametric Linear RegressionRidge Linear Regression: The Penalty TermFitting Ridge Regression ModelsPredicting Response Using Ridge Regression ModelsDetermining Ridge Regression Parameter, λThe Ridge Regression Model: Big Car DataThe Ridge Regression Model with Optimum λ: Big Car DataRegularized Parametric Linear Regression Model: LassoStepwise Parametric Linear RegressionFitting Stepwise Linear RegressionHow to Specify stepwiselm Model?Stepwise Linear Regression Model: Big Car DataCHW 4.1 Boston House PriceCHW 4.2 The Forest Fires DataCHW 4.3 The Parkinson’s Disease Telemonitoring DataCHW 4.4 The Car Fuel Economy Data
5 Neural Networks
IntroductionFeed-Forward Neural NetworksFeed-Forward Neural Network ClassificationFeed-Forward Neural Network RegressionNumeric Data: Dummy VariablesNeural Network Pattern Recognition (nprtool) ApplicationCommand-Based Feed-Forward Neural Network Classification: Heart DataNeural Network Regression (nftool)Command-Based Feed-Forward Neural Network Regression: Big Car DataTraining the Neural Network Regression Model Using fitrnet Function: Big Car DataFinding the Optimum Regularization Strength for Neural Network Using Cross-Validation: Big Car DataCustom Hyperparameter Optimization in Neural Network Regression: Big Car DataCHW 5.1 Mushroom Edibility DataCHW 5.2 1994 Adult Census Income DataCHW 5.3 Breast Cancer DiagnosisCHW 5.4 Small Car Data (Regression Case)CHW 5.5 Boston House Price

6 Pretrained Neural Networks: Transfer Learning
Deep Learning: Image NetworksData Stores in MATLABImage and Augmented Image DatastoresAccessing an Image FileRetraining: Transfer Learning for Image RecognitionConvolutional Neural Network (CNN) Layers: Channels and ActivationsConvolution 2-D Layer Features via ActivationsExtraction and Visualization of ActivationsA 2-D (or 2-D Grouped) Convolutional LayerFeatures Extraction for Machine LearningImage Features in Pretrained Convolutional Neural Networks (CNNs)Classification with Machine LearningFeature Extraction for Machine Learning: FlowersPattern Recognition Network GenerationMachine Learning Feature Extraction: SpectrogramsNetwork Object Prediction ExplainersOcclusion SensitivityimageLIME Features ExplainergradCAM Features ExplainerHCW 6.1 CNN Retraining for Round Worms Alive or Dead PredictionHCW 6.2 CNN Retraining for Food Images PredictionHCW 6.3 CNN Retraining for Merchandise Data PredictionHCW 6.4 CNN Retraining for Musical Instrument Spectrograms PredictionHCW 6.5 CNN Retraining for Fruit/Vegetable Varieties Prediction
7 A Convolutional Neural Network (CNN) Architecture and Training
A Simple CNN Architecture: The Land Satellite ImagesDisplaying Satellite ImagesTraining OptionsMini BatchesLearning RatesGradient ClippingAlgorithmsTraining a CNN for Landcover DatasetLayers and FiltersFilters in Convolution LayersViewing Filters: AlexNet FiltersValidation DataUsing shuffle FunctionImproving Network PerformanceTraining Algorithm OptionsTraining DataArchitectureImage Augmentation: The Flowers DatasetDirected Acyclic Graphs NetworksDeep Network Designer (DND)Semantic SegmentationAnalyze Training Data for Semantic SegmentationCreate a Semantic Segmentation NetworkTrain and Test the Semantic Segmentation NetworkHCW 7.1 CNN Creation for Round Worms Alive or Dead PredictionHCW 7.2 CNN Creation for Food Images PredictionHCW 7.3 CNN Creation for Merchandise Data PredictionHCW 7.4 CNN Creation for Musical Instrument Spectrograms PredictionHCW 7.5 CNN Creation for Chest X-ray PredictionHCW 7.6 Semantic Segmentation Network for CamVid Dataset
8 Regression Classification: Object Detection
Preparing Data for RegressionModification of CNN Architecture from Classification to RegressionRoot-Mean-Square ErrorAlexNet-Like CNN for Regression: Hand-Written Synthetic Digit ImagesA New CNN for Regression: Hand-Written Synthetic Digit ImagesDeep Network Designer (DND) for RegressionLoading Image DataGenerating Training DataCreating a Network ArchitectureImporting DataTraining the NetworkTest NetworkYOLO Object DetectorsObject Detection Using YOLO v4COCO-Based Creation of a Pretrained YOLO v4 Object DetectorFine-Tuning of a Pretrained YOLO v4 Object DetectorEvaluating an Object DetectorObject Detection Using R-CNN AlgorithmsR-CNNFast R-CNNFaster R-CNNTransfer Learning (Re-Training)R-CNN Creation and TrainingFast R-CNN Creation and TrainingFaster R-CNN Creation and TrainingevaluateDetectionPrecision Function for Precision MetricevaluateDetectionMissRate for Miss Rate MetricHCW 8.1 Testing yolov4ObjectDetector and fasterRCNN Object DetectorHCW 8.2 Creation of Two CNN-based yolov4ObjectDetectorsHCW 8.3 Creation of GoogleNet-Based Fast R-CNN Object DetectorHCW 8.4 Creation of a GoogleNet-Based Faster R-CNN Object DetectorHCW 8.5 Calculation of Average Precision and Miss Rate Using GoogleNet-Based Faster R-CNN Object DetectorHCW 8.6 Calculation of Average Precision and Miss Rate Using GoogleNet-Based yolov4 Object DetectorHCW 8.7 Faster RCNN-based Car Objects Prediction and Calculation of Average Precision for Training and Test Data
9 Recurrent Neural Network (RNN)
Long Short-Term Memory (LSTM) and BiLSTM NetworkTrain LSTM RNN Network for Sequence ClassificationImproving LSTM RNN PerformanceSequence LengthClassifying Categorical SequencesSequence-to-Sequence Regression Using Deep Learning: Turbo Fan DataClassify Text Data Using Deep Learning: Factory Equipment Failure Text Analysis – 1Classify Text Data Using Deep Learning: Factory Equipment Failure Text Analysis – 2Word-by-Word Text Generation Using Deep Learning – 1Word-by-Word Text Generation Using Deep Learning – 2Train Network for Time Series Forecasting Using Deep Network Designer (DND)Train Network with Numeric FeaturesHCW 9.1 Text Classification: Factory Equipment Failure Text AnalysisHCW 9.2 Text Classification: Sentiment Labeled Sentences Data SetHCW 9.3 Text Classification: Netflix Titles Data SetHCW 9.4 Text Regression: Video Game Titles Data SetHCW 9.5 Multivariate Classification: Mill Data SetHCW 9.6 Word-by-Word Text Generation Using Deep Learning
10 Image/Video-Based Apps
Image Labeler (IL) AppCreating ROI LabelsCreating Scene LabelsLabel Ground TruthExport Labeled Ground TruthVideo Labeler (VL) App: Ground Truth Data Creation, Training, and PredictionGround Truth Labeler (GTL) AppRunning/Walking Classification with Video Clips using LSTMExperiment Manager (EM) AppImage Batch Processor (IBP) AppHCW 10.1 Cat Dog Video Labeling, Training, and Prediction – 1HCW 10.2 Cat Dog Video Labeling, Training, and Prediction – 2HCW 10.3 EM Hyperparameters of CNN Retraining for Merchandise Data PredictionHCW 10.4 EM Hyperparameters of CNN Retraining for Round Worms Alive or Dead PredictionHCW 10.5 EM Hyperparameters of CNN Retraining for Food Images Prediction
Appendix A Useful MATLAB Functions
A.1 Data Transfer from an External Source into MATLABA.2 Data Import WizardA.3 Table OperationsA.4 Table Statistical AnalysisA.5 Access to Table Variables (Column Titles)A.6 Merging Tables with Mixed Columns and RowsA.7 Data PlottingA.8 Data NormalizationA.9 How to Scale Numeric Data Columns to Vary Between 0 and 1A.10 Random Split of a Matrix into a Training and Test SetA.11 Removal of NaN Values from a MatrixA.12 How to Calculate the Percent of Truly Judged Class Type Cases for a Binary Class ResponseA.13 Error Function m-fileA.14 Conversion of Categorical into Numeric Dummy MatrixA.15 evaluateFit2 FunctionA.16 showActivationsForChannel FunctionA.17 upsampLowRes FunctionA.18A preprocessData functionA.18B preprocessData2 functionA.19 processTurboFanDataTrain functionA.20 processTurboFanDataTest FunctionA.21 preprocessText FunctionA.22 documentGenerationDatastore FunctionA.23 subset Function for an Image Data Store Partition
Index
End User License Agreement

Content preview from Machine and Deep Learning Using MATLAB

7 A Convolutional Neural Network (CNN) Architecture and Training

Convolutional neural networks (CNNs) are a popular type of deep learning tool commonly used for various applications, particularly with images as inputs. However, they are also applicable to other domains like text, signals, and continuous responses. CNNs are inspired by the organization of simple and complex cells in the biological cortex, which respond to specific subregions (receptive regions) of a visual field. Similarly, CNNs are designed to activate based on subregions of an image rather than the entire image. In contrast, the fully connected (FC) layer is responsible for processing the entire image region. This means that neurons in a CNN focus only on their designated subregions and do not handle spatial areas beyond their assigned jurisdiction in the image.

Notice that such subregions might overlap, hence the neurons of a CNN produce spatially correlated outcomes, unlike other types of neural networks, where neurons do not share any connections and produce independent outcomes. Hence, in an FC layer, the number of parameters (weights) can tremendously increase as the size of the input increases. However, in a typical CNN the number of parameters will be reduced as a result of reduced number of connections, shared weights, and down-sampling.

The architecture of a CNN can vary depending on the types and number of layers to include. The types and number of layers to include depend on the intended application ...

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

MATLAB Machine Learning Recipes: A Problem-Solution Approach

Publisher Resources

ISBN: 9781394209088Purchase Link

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills

Machine and Deep Learning Using MATLAB

by Kamal I. M. Al-Malah

7 A Convolutional Neural Network (CNN) Architecture and Training

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.