Modern Business Analytics Chorzów

Deriving business value from analytics is a challenging process. Turning data into information requires a business analyst who is adept at multiple technologies including databases, programming tools, and commercial analytics tools. This practical guide shows programmers who understand analysis concepts how to build the skills necessary to achieve business value.Author Deanne Larson, data science practitioner and academic, helps you bridge the technical and business worlds to meet these requirements. You'll focus on developing these skills with R and Python using real-world examples. You'll also learn how to leverage methodologies for successful delivery. Learning methodology combined with open source tools is key to delivering successful business analytics and value.This book shows you how to:Apply business analytics methodologies to achieve successful resultsCleanse and transform data using R and PythonUse R and Python to complete exploratory data analysisCreate predictive models to solve business problems in R and PythonUse Python, R, and business analytics tools to handle large volumes of dataCommit code to GitHub to collaborate with data engineers and data scientistsMeasure success in business analytics Spis treści: Preface Who Should Read This Book Why I Wrote This Book Navigating This Book Conventions Used in This Book Using Code Examples OReilly Online Learning How to Contact Us Acknowledgments 1. The Role of Business Analyst and Analytics What Is the Role of a Business Analyst? Skills Responsibilities Types of Analysts Marketing analyst Financial analyst Functional analyst System analyst Data analyst Why Does a Business Analyst Need to Know Analytics? Data Explosion Business Context Analytics Descriptive Diagnostic Discovery Predictive Prescriptive Business Analyst Contributing to Analytics Value Business Problems Solved by Analytics Collaboration with Other Teams Skill Sets Used in Analytics Python and R Analytics Project Life Cycle Summary 2. Methodologies for the Business Analyst and Analytics Projects Business Understanding Determine Business Objectives Assess Situation Determine Goals Establish Approach and Plan Assessment of Tools and Techniques Data Exploration and Preparation Assess Data Content and Quality Select and Clean Data Construct and Integrate Data Produce Dataset for Model Development Modeling and Evaluation Select Analytics Technique Build and Assess Model Deployment Assess Model Performance Determine Assessment Intervals Model Operations Monitoring Models Life of a Model Retraining Summary 3. Introduction to R and Python R and Python Installation and Setup Options Why Learn R and Python? Learning Both at Once Versus One at a Time Pros and Cons of Different Learning Strategies R Installation Python Installation R and Python Scripting R Language Scripting Python Language Scripting Object-Oriented Concepts Structure of OOP Class Object Methods and attributes Principles of OOP Encapsulation Abstraction Inheritance Polymorphism R and Python Data Types R Data Types R Structures Python Data Types Python Data Structures Interaction with Relational Databases Why Relational Databases? R Connection to Relational Databases Examples of R and Relational Databases SQLite Python Connection to Relational Databases Examples of Python and Relational Databases Summary 4. Statistical Analysis with R and Python Example Analytical Projects Telecom Churn A/B Testing Marketing Campaigns Financial Forecasting Healthcare Diagnosis Starting with the Problem Statement Getting to the Analytical Problem Classification Regression What Do We Want to Measure? Analysis Approaches EDA Unsupervised Learning Statistical Analysis for Regression Analysis for Classification Role of Hypothesis Testing Visualization in Analytics Visualization in R and Python to Support EDA Regression Visualization Scatter plots Box plots Density plots Heatmaps Classification Visualization Bar plots Parallel coordinates plot Violin plots Contour plots Summary 5. Exploratory Data Analysis with R and Python Data Quality Data Quality Characteristics Data Profiling Clustering and Unsupervised Learning Purpose of Unsupervised Learning Example of Clustering Impacting Supervised Learning K-Means Clustering Hierarchical Clustering Other Unsupervised Methods Used in EDA Identifying Outliers Outliers in Regression Outliers in Classification Data Preparation for Modeling Sampling Random sampling Stratified sampling Systematic sampling Cluster sampling Bootstrap sampling Oversampling and undersampling Importance of sampling in model building Training and Testing Data Transformation Data formatting One-hot encoding Binning Derived attributes Scaling, normalization, and standardization Data Manipulation Selecting and Reducing Features Feature Selection Filter methods Iterative methods Wrapper methods Embedded methods Feature Reduction Techniques Feature reduction for regression PCA LDA Feature reduction for classification Summary 6. Application and Evaluation of Modeling in R and Python Modeling Steps Model Selection and Training Model Evaluation Model Optimization Model Deployment Model Monitoring and Maintenance Selecting the Right Algorithm Regression Common Use Cases Linear Regression Equation Linear Regression in R Linear Regression in Python Coefficients and statistics Residual and model diagnostics Linear Regression Use Case Other Types of Regression Polynomial regression Multivariate regression Time series regression LASSO regression Ridge regression Elastic net Challenges with Regression Models Other Algorithms for Regression Decision Trees for Regression Distinguishing regression trees from classification trees Linear Regression Evaluation Model evaluation in R Model evaluation in Python Classification Common Use Cases Classification Algorithms Classification in R Classification in Python Classification Use Case: Telecom Churn Python example Classification Evaluation Metrics Confusion matrix Model evaluation in R Model evaluation in Python Calculating classification metrics in Python Classification Use Case Evaluation Summary 7. Modeling and Algorithm Choice Algorithms Algorithm Criteria Problem Type Classification problems Regression problems Clustering problems Dimensionality reduction Interpretable Models Linear models Decision trees Ensemble models Generalized additive models Prediction Accuracy Complexity and capacity Ensemble methods SVMs Feature engineering Training Speed Algorithmic efficiency Data size and quality Parallelization and distributed computing Algorithm selection based on problem complexity Early stopping Prediction Speed Model complexity Dimensionality reduction Model training and serving architecture Algorithm computational speed Model pruning and quantization Hyperparameter Tuning Algorithm complexity and hyperparameter space Automated hyperparameter tuning tools Computational resources Tuning versus performance trade-off Sensitivity analysis Cross-validation strategy Example of hyperparameter tuning Working with a Small Dataset Working with a Large Dataset Feature Interaction Decision trees Deep learning models Kernel methods Regularization techniques Data Characteristics Dimensionality Feature type Class distribution (balanced versus imbalanced) Data quality and missing values Underlying data distribution Example: Selecting the Right Algorithm Choosing the Right Algorithm to Predict Sales Step 1: Understanding the problem type Step 2: Model interpretation Step 3: Model accuracy Step 4: Training speed Step 5: Prediction speed Step 6: Parameter tuning Step 7: Size of dataset Step 8: Feature interaction Evaluating the Criteria Decision and Implementation Summary 8. Model Operations Overview of Model Operations Model Operations Processes Model Scoring Model Scoring in R: Using Shiny Apps for Real-Time Scoring Model Scoring in Python: Deploying Models with Streamlit Model Monitoring Key Metrics and Indicators for Model Performance Monitoring Techniques for Automated Model Monitoring Implementation in R: Building dashboards with shinydashboard Implementation in Python: Utilizing visualization libraries like Matplotlib and Seaborn Step 1: Install necessary packages Step 2: Sample Python code Model Retraining Triggering Events for Model Retraining Techniques for Automated Model Retraining Implementation in R: Using cron Jobs for Scheduled Retraining Implementation in Python: Leveraging Tools Like Airflow for Workflow Management Generating Reports Content and Structure of Final Reports Techniques for Automated Report Generation Implementation in R: Generating Reports with R Markdown and knitr Step 1: Document setup Step 2: Code integration Step 3: Knitting process Step 4: Automation Implementation in Python: Creating Reports with Jupyter Notebooks and nbconvert Step 1: Notebook development Step 2: Markdown cells Step 3: nbconvert conversion Step 4: Automation Version Control and Model Reproducibility Collaboration and Documentation Practices ModelOps Use Cases Retail Sales Forecasting: Automation of Scoring and Monitoring Fraud Detection: Dynamic Model Retraining and Reporting Customer Churn Prediction: Scheduled Model Retraining and Final Report Generation Integration with Existing Systems and Infrastructure Future Direction of MLOps Summary 9. Advanced Visualization Advanced Visualization with R Shiny What Is R Shiny? Key Features and Capabilities of R Shiny Interactive web applications Reactivity Accessibility Extensibility Community and support Setting Up Your Environment Building Your First Shiny App Structure of a Shiny app: UI and server components A simple example app explained Interactive graphics in Shiny Using Plotly and ggplot2 for dynamic plots ggplot2 Plotly Advanced UI Development Customizing appearance with HTML and CSS HTML (HyperText markup language) CSS Relationship and usage Using shinydashboard for creating dashboards Shiny widgets and extensions Example: Creating a Dashboard to Monitor Real-Time Sales Learning Python Visualization Overview of Visualization in Python Common Libraries: Matplotlib, Seaborn, Plotly, and Dash Matplotlib: Foundations of Visualization in Python Customizing Plots with Styles and Colors Statistical Plots: Scatter Plots, Heatmaps, Violin Plots Interactive Plots with Plotly 3D Plotting with Matplotlib and Plotly Geospatial Data Visualization Dashboard Creation: Use Plotly Dash Case Study: Using Python for an Advanced Visualization Project Choosing Between R Shiny and Python Visualization Summary 10. Working with Modern Data Types in Analytics Semistructured Data (JSON) Using Python for JSON Data Loading and parsing JSON data Extracting data from nested JSON structures Transforming JSON data into Pandas DataFrames Cleaning and normalizing JSON data Analyzing JSON data with Python Using R for JSON Data Loading and parsing JSON data Extracting data from nested JSON structures Transforming JSON data into data frames with R Cleaning and normalizing JSON data Analyzing JSON data with R Social Media Data Types of Social Media Data Posts Comments Likes and reactions Shares and reposts Direct messages Using Python for Social Media Data Analysis X data extraction using Tweepy Setting up the X API Fetching posts using Tweepy Cleaning and processing X posts with Tweepy Facebook and data extraction using Facebook Graph API Setting up Facebook Graph API Data extraction using Facebook Graph API Sentiment analysis on social media data in Python Performing sentiment analysis with TextBlob or NLTK libraries Visualizing sentiment trends with Matplotlib or Seaborn libraries Using R for Social Media Data Analysis X data extraction using rtweet Fetching posts using rtweet Cleaning and processing tweet data using rtweet Analyzing and visualizing post data Facebook and Instagram data extraction Installing necessary packages Fetching Facebook data Fetching Instagram data Cleaning and processing data Sentiment analysis on social media data in R Image Data Image Processing in Python Image Processing in R Video Data Using Python for Video Data Extracting frames and saving them as images Analyzing video content Basic image processing Object detection using pretrained models Storing and visualizing video data insights Creating visual summaries of video analysis Using R for Video Data Extracting frames from video Extracting frames and saving them as images Analyzing video content Storing and visualizing video data insights Creating visual summaries of video analysis Summary 11. Measuring Business Value from Analytics and the Role of AI What Is Business Value in Analytics? Strategic Impact Operational Efficiency Customer Satisfaction and Loyalty Metrics and KPIs for Measuring Business Value Financial Metrics Operational Metrics Customer Metrics Aligning Metrics with Organizational Goals and Objectives Leveraging Metrics to Demonstrate Value Metrics and KPIs in Practice Business Case Examples of Value for Analytics Step 1: Problem Definition and Setting Measurable Outcomes Step 2: Identifying Metrics to Measure Success and Failure Step 3: Implementing Analytics Solutions Step 4: Measuring and Demonstrating Value Step 5: Reporting and Continuous Improvement AI and Generative AI in Business Analytics Introduction to Generative AI Applications in Product Design Applications in Content Creation Applications in Marketing Enhancing Customer Experience Improving Operational Efficiency Future Prospects and Challenges Use Cases for AI and Generative AI in Business Analytics Use Case 1: AI-Driven Customer Insights and Recommendations Use Case 2: Generative AI in Content Creation Use Case 3: AI-Powered Supply Chain Optimization Use Case 4: Enhancing Decision Making with AI Use Case 5: AI in Healthcare Analytics Use Case 6: Generative AI for Personalized Customer Experiences Use Case 7: AI in Retail Analytics Addressing Factual Inconsistencies and Human-AI Collaboration Future Prospects Challenges and Considerations Integration Challenges and Scalability in Deploying AI Solutions Mitigating Biases and Ensuring Fairness in AI-Driven Decisions Technical and Organizational Challenges in AI Deployment Cost and Resource Considerations Future-Proofing AI Investments Summary Index