Data Science for Decision Makers: Enhance your leadership skills with data science and AI expertise

Data Science for Decision Makers
Contributors
About the author
About the reviewer
Preface
   Who this book is for
   What this book covers
   Conventions used
   Get in touch
   Share Your Thoughts
   Download a free PDF copy of this book
Part 1: Understanding Data Science and Its Foundations
1
Introducing Data Science
   Data science, AI, and ML – what’s the difference?
      The mathematical and statistical underpinnings of data science
   Statistics and data science
      What is statistics?
   Descriptive and inferential statistics
      Sampling strategies
   Probability
      Probability distribution
      Conditional probability
   Describing our samples
      Measures of central tendency
      Measures of dispersion
      Degrees of freedom
      Correlation, causation, and covariance
      The shape of data
   Probability distributions
      Discrete probability distributions
      Continuous probability distributions
   Summary
2
Characterizing and Collecting Data
   What are the key criteria to consider when evaluating datasets?
      Data quantity
      Data velocity
      Data variety
      Data quality
   First-, second-, and third-party data
      First-party data – the treasure trove within
      Second-party data – building bridges through collaboration
      Third-party data – broadening horizons with external expertise
   Structured, unstructured, and semi-structured data
      Structured data
      Unstructured data
      Semi-structured data
   Methods for collecting data
   Storing and processing data
   Cloud, on-premises, and hybrid solutions – navigating the data storage and analysis landscape
      Cloud computing – scalable services in the cloud
      On-premises – maintaining control within your walls
      Hybrid – the best of both worlds?
   Data processing
   Summary
3
Exploratory Data Analysis
   Getting started with Google Colab
      What is Google Colab?
      A step-by-step guide to setting up Google Colab
   Understanding the data you have
   EDA techniques and tools
      Descriptive statistics
      Data visualization
      Histograms
      Density curves
      Boxplots
      Heatmaps
      Dimensionality reduction
      Correlation analysis
      Outlier detection
   Summary
4
The Significance of Significance
   The idea of testing hypotheses
      What is a hypothesis?
      How does hypothesis testing work?
      Formulating null and alternative hypotheses
      Determining the significance level
      Understanding errors
      Getting to grips with p-values
   Significance tests for a population proportion – making informed decisions about proportions
      The z-test – comparing a sample proportion to a population proportion
      Z-test example made easy
   Significance tests for a population average (mean)
      Writing hypotheses for a significance test about a mean
      Conditions for a t-test about a mean
      When to use z or t statistics in significance tests
      Example – calculating the t-statistic for a test about a mean
      Using a table to estimate the p-value from the t-statistic
      Comparing the p-value from the t-statistic to the significance level
      One-tailed and two-tailed tests
   Walking through a case study
   Summary
5
Understanding Regression
   How can I benefit from understanding regression?
   Introduction to trend lines
   Fitting a trend line to data
   Estimating the line of best fit
   Calculating the equations of the lines of best fit
   Interpreting the slope of a regression line
   Interpreting the intercept of a regression line
   Understanding residuals
   Evaluating the goodness of fit in least-squares regression
   Summary
Part 2: Machine Learning – Concepts, Applications, and Pitfalls
6
Introducing Machine Learning
   From statistics to machine learning
      What is machine learning?
      How does machine learning relate to statistics?
   Why is machine learning important?
      Customer personalization and segmentation
      Fraud detection and security
      Supply chain and inventory optimization
      Predictive maintenance
      Healthcare diagnostics and treatment
   The different types of machine learning
      Supervised learning
      Unsupervised learning
      Semi-supervised learning
      Reinforcement learning
      Transfer learning
   Popular machine learning algorithms
      Linear regression
      Logistic regression
      Decision trees
      Random forests
      Support vector machines
      k-nearest neighbors
      Neural networks
   The machine learning process
      Training a supervised machine learning model
      Validation of a supervised machine learning model
      Testing a supervised machine learning model
      Evaluating machine learning models
   Risks and limitations of machine learning
      Overfitting and underfitting
      Bias and variance
      Balanced dataset
      Models are approximations of reality
   Machine learning on unstructured data
      Natural language processing (NLP)
      Computer vision
   Deep learning and artificial intelligence
      Artificial intelligence
      Deep learning
   Summary
7
Supervised Machine Learning
   Defining supervised learning
      Applications of supervised learning
      The two types of supervised learning
      Key factors in supervised learning
   Steps within supervised learning
      Data preparation – laying the foundation
      Algorithm selection – choosing the right tool
      Model training – learning from data
      Model evaluation – assessing performance
      Prediction and deployment – putting the model to work
   Characteristics of regression and classification algorithms
      Regression algorithms
      Classification algorithms
      Key considerations in supervised learning
      Evaluation metrics
   Applications of supervised learning
      Consumer goods
      Retail
      Manufacturing
   Summary
8
Unsupervised Machine Learning
   Defining UL
      Practical examples of UL
   Steps in UL
      Step 1 – Data collection
      Step 2 – Data preprocessing
      Step 3 – Choosing the right model
      Step 4 – Training the model
      Step 5 – Interpretation and evaluation
      In summary
   Clustering – unveiling hidden patterns in your data
      What is clustering?
      How does clustering work?
      k-means clustering
      Practical applications of clustering
      Evaluation metrics for clustering
      In summary
   Association rule learning
      What is association rule learning?
      The Apriori algorithm – a practical example
      Evaluation metrics
      In summary
   Applications of UL
      Market segmentation
      Anomaly detection
      Feature extraction
   Summary
9
Interpreting and Evaluating Machine Learning Models
   How do I know whether this model will be accurate?
      Evaluating on test (holdout) data
   Understanding evaluation metrics
      Evaluating regression models
      R-squared
      Root mean squared error
      Mean absolute error
      When and how to use each metric
      Practical evaluation strategies
      Summarizing the evaluation of regression models
   Evaluating classification models
      Classification model evaluation metrics
      Precision, recall, and F1-Score
      Recall
      F1-score
   Methods for explaining machine learning models
      Making sense of regression models – the power of coefficients
      Decoding classification models – unveiling feature importance
      Beyond specific models – universal insights using SHAP values
   Summary
10
Common Pitfalls in Machine Learning
   Understanding the complexity
   Dirty data, damaged models – how data quantity and quality impact ML
      The importance of adequate training data
      Dealing with poor data quality
      Conclusion
   Overcoming overfitting and underfitting
      Navigating training-serving skew and model drift
      Ensuring fairness
   Mastering overfitting and underfitting for optimal model performance
      Overfitting – when your model is too specific
      Underfitting – when your model is too simplistic
      Spotting the problem
      Conclusion
   Training-serving skew and model drift
      Training-serving skew
      Model drift
      Key takeaways
   Bias and fairness
      Understanding bias
      Understanding fairness
      Mitigating bias and ensuring fairness
      Key takeaways
   Summary
Part 3: Leading Successful Data Science Projects and Teams
11
The Structure of a Data Science Project
   The various types of data science projects
      Data products
      Reports and analytics
      Research and methodology
   The stages of a data product
      Identifying use cases
      Evaluating use cases
      Planning the data product
   Developing a data product
      Data preparation and exploratory analysis
      Model design and development
      Evaluation and testing
   Deploying and monitoring a data product
   General best practices for data product development
   Evaluating impact
      Predictive maintenance in manufacturing
      Fraud detection in banking
      Customer churn prediction in telecom
      Demand forecasting in retail
      Personalized recommendations in e-commerce
      Predictive maintenance in energy
      Workforce optimization in quick service restaurants
      Chatbot-assisted customer support
   Summary
12
The Data Science Team
   Assembling your data science team – key roles and considerations
      Data scientists
      Machine learning engineers
      Data engineers
      MLOps engineers
      Analytics engineers
      Software engineers (full stack, frontend, backend)
      Product managers
      Business analysts
      Data storytellers/visualization experts
      Considerations when assembling your team
      Data science teams within larger organizations
   The hub and spoke model
      What is the hub and spoke model?
      Practical applications of the hub and spoke model
      Building a hub and spoke model
   The art of recruitment
      Where to find technical talent
   How high-performing data science teams operate
      Cross-functional collaboration is essential
      Diversity of perspectives drives innovation
      Start with the right problem to solve
      Invest in tooling, infrastructure, and workflow
      Continuous adaption and learning are a must
      Focus ruthlessly on outcomes over activity
   Summary
13
Managing the Data Science Team
   Day-to-day management of a data science team
      Enabling rapid experimentation and innovation
      Managing inherent uncertainty
      Balancing research and application
      Communicating effectively in data science and artificial intelligence
      Fostering a culture of curiosity and continuous learning
      Embracing peer review and collaboration
   Common challenges in managing a data science team
      Challenge 1 – recruiting and retaining top talent
      Challenge 2 – aligning projects with business goals
      Challenge 3 – managing inherent uncertainty
      Challenge 4 – scaling and operationalizing models
      Challenge 5 – deploying robust, reliable, fair models ethically
   Empowering and motivating your data science team
      Working with other teams and external stakeholders and empowering them to use data
   Summary
14
Continuing Your Journey as a Data Science Leader
   Navigating the landscape of emerging technologies
   Specializing in an industry
   Specializing in a field
   Embracing continuous learning
      Online courses
      Cloud certifications
      Technical tutorials and documentation
      Learning plan framework
   Staying up to date with current DS/ML/AI news and trends
   Promoting data-driven thinking within your organization
      Host internal learning sessions
      Collaborate on cross-functional projects
      Share success stories and lessons learned
      Mentor and upskill colleagues
      Establish a data science community of practice
   Networking beyond your organization
      Attend industry conferences and events
      Join online communities and forums
      Engage with local meetups and user groups
      Collaborate on side projects or research
      Offer mentorship or seek mentors
   Summary
Index
   Why subscribe?
Other Books You May Enjoy
   Packt is searching for authors like you
   Share Your Thoughts
   Download a free PDF copy of this book