Applied Deep Learning with Python(英文版).pdf

如果你是一个Python程序员，走进数据科学的世界，这是正确的方式开始。对于希望使用TensorFlow和Keras的有经验的开发人员、分析人员或数据科学家来说，它也是理想的选择。我们假设您熟悉Python、Web应用程序开发、Docker命令以及线性代数、概率和统计的概念。

Table of ContentsTitle PageCopyright and CreditsApplied Deep Learning with PythonPackt UpsellWhy subsTo get the most out of this bookDownload the example code filesConventions usedGet in touchReviews1. Jupyter FundamentalsBasic Functionality and FeaturesWhat is a Jupyter Notebook and Why is it Useful?Navigating the PlatformIntroducing Jupyter NotebooksJupyter FeaturesExploring some of Jupyter's most useful featuresConverting a Jupyter Notebook to a Python ScriptPython LibrariesImport the external libraries and set up the plotting environmentOur First Analysis - The Boston Housing DatasetLoading the Data into Jupyter Using a Pandas DataFrameLoad the Boston housing datasetData ExplorationExplore the Boston housing datasetIntroduction to Predictive Analytics with Jupyter NotebooksLinear models with Seaborn and scikit-learnActivity:Building a Third-Order Polynomial ModelLinear models with Seaborn and scikit-learnUsing Categorical Features for Segmentation AnalysisCreate categorical filelds from continuous variables and make segmented visualizationsSummary2. Data Cleaning and Advanced Machine LearningPreparing to Train a Predictive ModelDetermining a Plan for Predictive AnalyticsPreprocessing Data for Machine LearningExploring data preprocessing tools and methodsActivity:Preparing to Train a Predictive Model for the Employee-RetentionProblemTraining Classification ModelsIntroduction to Classification AlgorithmsTraining two-feature classification models with scikitlearnThe plot_decision_regions FunctionTraining k-nearest neighbors for our modelTraining a Random ForestAssessing Models with k-Fold Cross-Validation and Validation CurvesUsing k-fold cross validation and validation curves in Python with scikit-learnDimensionality Reduction TechniquesTraining a predictive model for the employee retention problemSummary3. Web Scraping and Interactive VisualizationsScraping Web Page DataIntroduction to HTTP RequestsMaking HTTP Requests in the Jupyter NotebookHandling HTTP requests with Python in a Jupyter NotebookParsing HTML in the Jupyter NotebookParsing HTML with Python in a Jupyter NotebookActivity:Web Scraping with Jupyter NotebooksInteractive VisualizationsBuilding a DataFrame to Store and Organize DataBuilding and merging Pandas DataFramesIntroduction to BokehIntroduction to interactive visualizations with BokehActivity:Exploring Data with Interactive VisualizationsSummary4. Introduction to Neural Networks and DeepLearning What are Neural Networks?Successful ApplicationsWhy Do Neural Networks Work So Well?Representation LearningFunction ApproximationLimitations of Deep LearningInherent Bias and Ethical ConsiderationsCommon Components and Operations of Neural NetworksConfiguring a Deep Learning EnvironmentSoftware Components for Deep LearningPython 3TensorFlowKerasTensorBoardJupyter Notebooks, Pandas, and NumPyActivity:Verifying Software ComponentsExploring a Trained Neural NetworkMNIST DatasetTraining a Neural Network with TensorFlowTraining a Neural NetworkTesting Network Performance with Unseen DataActivity: Exploring a Trained Neural NetworkSummary5. Model ArchitectureChoosing the Right Model ArchitectureCommon ArchitecturesConvolutional Neural NetworksRecurrent Neural NetworksGenerative Adversarial NetworksDeep Reinforcement LearningData NormalizationZ-scorePoint-Relative NormalizationMaximum and Minimum NormalizationStructuring Your ProblemActivity:Exploring the Bitcoin Dataset and Preparing Data for ModelUsing Keras as a TensorFlow InterfaceModel ComponentsActivity:Creating a TensorFlow Model Using KerasFrom Data Preparation to ModelingTraining a Neural NetworkReshaping Time-Series DataMaking PredictionsOverfittingActivity:Assembling a Deep Learning SystemSummary6. Model Evaluation and OptimizationModel EvaluationProblem CategoriesLoss Functions, Accuracy, and Error RatesDifferent Loss Functions, Same ArchitectureUsing TensorBoardImplementing Model Evaluation MetricsEvaluating the Bitcoin ModelOverfittingModel PredictionsInterpreting PredictionsActivity:Creating an Active Training EnvironmentHyperparameter OptimizationLayers and Nodes - Adding More LayersAdding More NodesLayers and Nodes - ImplementationEpochsEpochs - ImplementationActivation FunctionsLinear (Identity)Hyperbolic Tangent (Tanh)Rectifid Linear UnitActivation Functions - ImplementationRegularization StrategiesL2 RegularizationDropoutRegularization Strategies – ImplementationOptimization ResultsActivity:Optimizing a Deep Learning ModelSummary7. ProductizationHandling New DataSeparating Data and ModelData ComponentModel ComponentDealing with New DataRe-Training an Old ModelTraining a New ModelActivity:Dealing with New DataDeploying a Model as a Web ApplicationApplication Architecture and TechnologiesDeploying and Using CryptonicActivity:Deploying a Deep Learning ApplicationSummaryOther Books You May EnjoyLeave a review - let other readers know what you think