Smarter Decisions – The Intersection of Internet of Things and Decision Science

Book description

Enter the world of Internet of Things with the power of data science with this highly practical, engaging book

About This Book

  • Explore real-world use cases from the Internet of Things (IoT) domain using decision science with this easy-to-follow, practical book
  • Learn to make smarter decisions on top of your IoT solutions so that your IoT is smart in a real sense
  • This highly practical, example-rich guide fills the gap between your knowledge of data science and IoT

Who This Book Is For

If you have a basic programming experience with R and want to solve business use cases in IoT using decision science then this book is for you. Even if your're a non-technical manager anchoring IoT projects, you can skip the code and still benefit from the book.

What You Will Learn

  • Explore decision science with respect to IoT
  • Get to know the end to end analytics stack ? Descriptive + Inquisitive + Predictive + Prescriptive
  • Solve problems in IoT connected assets and connected operations
  • Design and solve real-life IoT business use cases using cutting edge machine learning techniques
  • Synthesize and assimilate results to form the perfect story for a business
  • Master the art of problem solving when IoT meets decision science using a variety of statistical and machine learning techniques along with hands on tasks in R

In Detail

With an increasing number of devices getting connected to the Internet, massive amounts of data are being generated that can be used for analysis. This book helps you to understand Internet of Things in depth and decision science, and solve business use cases. With IoT, the frequency and impact of the problem is huge. Addressing a problem with such a huge impact requires a very structured approach.

The entire journey of addressing the problem by defining it, designing the solution, and executing it using decision science is articulated in this book through engaging and easy-to-understand business use cases. You will get a detailed understanding of IoT, decision science, and the art of solving a business problem in IoT through decision science.

By the end of this book, you'll have an understanding of the complex aspects of decision making in IoT and will be able to take that knowledge with you onto whatever project calls for it

Style and approach

This scenario-based tutorial approaches the topic systematically, allowing you to build upon what you learned in previous chapters.

Publisher resources

Download Example Code

Table of contents

  1. Smarter Decisions – The Intersection of Internet of Things and Decision Science
    1. Smarter Decisions – The Intersection of Internet of Things and Decision Science
    2. Credits
    3. About the Author
    4. About the Reviewer
    5. eBooks, discount offers, and more
      1. Why subscribe?
    6. Preface
      1. What this book covers
      2. What you need for this book
      3. Who this book is for
      4. Sections
        1. Getting ready
        2. How to do it…
        3. How it works…
        4. There's more…
        5. See also
      5. Conventions
      6. Reader feedback
      7. Customer support
        1. Downloading the example code
        2. Errata
        3. Piracy
        4. Questions
    7. 1. IoT and Decision Science
      1. Understanding the IoT
        1. IoT in a real-life scenario
      2. Demystifying M2M, IoT, IIoT, and IoE
      3. Digging deeper into the logical stack of IoT
        1. People
        2. Processes
          1. Technology
            1. Software
            2. Protocol
            3. Infrastructure
          2. Business processes
        3. Things
        4. Data
      4. The problem life cycle
      5. The problem landscape
      6. The art of problem solving
        1. The interdisciplinary approach
        2. The problem universe
      7. The problem solving framework
      8. Summary
    8. 2. Studying the IoT Problem Universe and Designing a Use Case
      1. Connected assets & connected operations
        1. The journey of connected things to smart things
        2. Connected assets - A real life scenario
        3. Connected operations – The next revolution
          1. What is Industry 4.0?
      2. Defining the business use case
        1. Defining the problem
        2. Researching and gathering context
          1. Gathering context - examining the type of problem
          2. Gathering context - research and gather context
          3. Research outcome
            1. How is detergent manufactured?
            2. What are the common issues that arise in the detergent manufacturing process?
            3. What kind of machinery is used for the detergent manufacturing process?
            4. What do we need to know more about the company, its production environment, and operations?
        3. Prioritize and structure hypotheses based on the availability of data
        4. Validating and Improving the hypotheses (iterate over #2 and #3)
        5. Assimilate results and render the story
      3. Sensing the associated latent problems
      4. Designing the heuristic driven hypotheses matrix (HDH)
      5. Summary
    9. 3. The What and Why - Using Exploratory Decision Science for IoT
      1. Identifying gold mines in data for decision making
        1. Examining data sources for the hypotheses
        2. Data surfacing for problem solving
          1. End product related information
          2. Manufacturing environment information
          3. Raw material data
          4. Operational data
          5. Summarizing the data surfacing activity
        3. Feature exploration
        4. Understanding the data landscape
          1. Domain context for the data
      2. Exploring each dimension of the IoT Ecosystem through data (Univariates)
        1. What does the data say?
        2. Exploring Previous Product...
        3. Summarizing this section
      3. Studying relationships
        1. So what is correlation?
        2. Exploring Stage 1 dimensions
          1. Revisiting the DDH matrix
      4. Exploratory data analysis
        1. So how do we validate our findings?
        2. So how does hypothesis testing work?
        3. Validating hypotheses - category 1
        4. How does the chi-squared test work in a nutshell?
        5. Validating hypotheses - category 2
          1. What does a Type 1 error mean?
          2. So what is ANOVA?
        6. Validating hypotheses - category 3
          1. So what is regression?
        7. Hypotheses - category 3
        8. Summarizing Exploratory Data Analysis phase
      5. Root Cause Analysis
        1. Synthesizing results
        2. Visualizing insights
        3. Stitching the Story together
        4. Conclusion
          1. Production Quantity
          2. Raw material quality parameters
          3. Resources/Machinery used in Stage 3
          4. Assembly Line
      6. Summary
    10. 4. Experimenting Predictive Analytics for IoT
      1. Resurfacing the problem - What's next?
      2. Linear regression - predicting a continuous outcome
        1. Prelude
        2. Solving the prediction problem
          1. So what is linear regression?
        3. Interpreting the regression outputs
          1. F statistic
          2. Estimate/coefficients
          3. Standard error, t-value, and p value
        4. Residuals, multiple R squared, residual standard error and adjusted R squared
          1. What is the adjusted R-squared value?
        5. Improving the predictive model
          1. Let's define our approach
          2. How will we go about it?
          3. Let's being modeling
          4. So how do we move ahead?
          5. The important points to ponder are as follows:
          6. What should we take care of?
          7. So what next?
      3. Decision trees
        1. Understanding decision trees
          1. So what is a decision tree?
          2. How does a decision tree work?
          3. What are different types of decision trees?
          4. So how is a decision tree built and how does it work?
          5. How to select the root node?
          6. How are the decision nodes ordered/chosen?
          7. How different is the process for classification and regression?
        2. Predictive modeling with decision trees
          1. So how do we approach?
          2. So what do we do to improve the results?
          3. So, what next? Do we try another modeling technique that could give us more powerful results?
      4. Logistic Regression - Predicting a categorical outcome
        1. So what is logistic regression?
        2. So how does the logistic regression work?
          1. How do we assess the goodness of fit or accuracy of the model?
          2. Too many new terms?
        3. Recap to the model interpretation
        4. Improving the classification model
          1. Let's define our approach
          2. How do we go about it?
          3. Let's begin modeling
          4. So how do we move ahead?
          5. Adding interaction terms
          6. What can be done to improve this?
          7. What just happened?
          8. What can be done to improve the TNR and overall accuracy while keeping the TPR intact?
      5. Summary
    11. 5. Enhancing Predictive Analytics with Machine Learning for IoT
      1. A Brief Introduction to Machine Learning
        1. What exactly is ensemble modeling?
        2. Why should we choose ensemble models?
        3. So how does an ensemble model actually work?
          1. What are the different ensemble learning techniques?
          2. Quick Recap - Where were we previously?
      2. Ensemble modeling - random forest
        1. What is random forest?
        2. How do we build random forests in R?
          1. What are these new parameters?
          2. Mtry
          3. Building a more tuned version of the random forest model
          4. How?
          5. Can we improve this further?
          6. What can we do to achieve this?
      3. Ensemble modeling - XGBoost
        1. What is different in XgBoost?
          1. Are we really getting good results?
          2. What next?
          3. A cautionary note
      4. Neural Networks and Deep Learning
        1. So what is so cool about neural networks and deep learning?
          1. What is a neural network?
          2. So what is deep learning?
          3. So what problems can neural networks and deep learning solve?
          4. So how does a neural network work?
          5. Neurons
          6. Edges
          7. Activation function
          8. Learning
          9. So what are the different types of neural networks?
          10. How do we go about modeling using a neural network or deep learning technique?
          11. What next?
          12. What have we achieved till now?
      5. Packaging our results
        1. A quick recap
        2. Results from our predictive modeling exercise
        3. Few points to note
      6. Summary
    12. 6. Fast track Decision Science with IoT
      1. Setting context for the problem
        1. The real problem
        2. What next?
      2. Defining the problem and designing the approach
        1. Building the SCQ: Situation - Complication - Question
        2. Research
          1. How does a solar panel ecosystem work?
          2. Functioning
          3. What are the different kinds of solar panel installations?
          4. What challenges are faced in operations supported by solar panels?
        3. Domain context
        4. Designing the approach
        5. Studying the data landscape
      3. Exploratory Data Analysis and Feature Engineering
        1. So how does the consumption fare in comparison with the generation?
        2. Battery
        3. Load
        4. Inverter
        5. Assimilate learnings from the data exploration exercise
        6. Let's assimilate all our findings and learnings in brief
        7. Solving the problem
        8. Feature engineering
      4. Building predictive model for the use case
        1. Building a random forest model
      5. Packaging the solution
      6. Summary
    13. 7. Prescriptive Science and Decision Making
      1. Using a layered approach and test control methods to outlive business disasters
        1. What is prescriptive analytics?
          1. What happened?
          2. Why and how did it happen?
          3. When will it happen (again)?
          4. So what, now what?
        2. Solving a prescriptive analytics use case
          1. Context for the use case
          2. Descriptive analytics - what happened?
          3. Inquisitive analytics - why and how did it happen?
          4. Predictive analytics – when will it happen?
          5. The inception of prescriptive analytics
          6. Getting deeper with prescriptive analytics
        3. Solving the use case the prescriptive way
          1. Test and control analysis
          2. Implementing Test & Control Analysis in Prescriptive Analytics
          3. Improving IVR operations to increase the call completion rate
          4. Reducing the repeat calls
          5. Staff training for increasing first call resolution rate
          6. Tying back results to data-driven and heuristic-driven hypotheses
      2. Connecting the dots in the problem universe
      3. Story boarding - Making sense of the interconnected problems in the problem universe
        1. Step 1 - Immediate
        2. Step 2 - Future
      4. Implementing the solution
      5. Summary
    14. 8. Disruptions in IoT
      1. Edge/fog computing
        1. Exploring the fog computing model
      2. Cognitive Computing - Disrupting intelligence from unstructured data
        1. So how does cognitive computing work?
        2. Where do we see the use of cognitive computing?
        3. The story
        4. The bigger question is, how does all of this happen?
      3. Next generation robotics and genomics
        1. Robotics – A bright future with IoT, Machine Learning, Edge & Cognitive Computing
        2. Genomics
        3. So how does genomics relate to IoT?
      4. Autonomous cars
        1. Vision and inspiration
        2. So how does an autonomous car work?
        3. Wait, what are we missing?
        4. Vehicle - to - environment
        5. Vehicle - to - vehicle
        6. Vehicle - to - infrastructure
        7. The future of autonomous cars
      5. Privacy and security in IoT
        1. Vulnerability
        2. Integrity
        3. Privacy
        4. Software infrastructure
        5. Hardware infrastructure
        6. The protocol infrastructure
      6. Summary
    15. 9. A Promising Future with IoT
      1. The IoT Business model - Asset or Device as a Service
        1. The motivation
        2. Real life use case for Asset as a Service model
        3. How does it help business?
          1. Best case scenario
          2. Worst case scenario
          3. Neutral case
          4. Conclusion
        4. Leveraging Decision Science to empower the Asset as a Service model
      2. Smartwatch – A booster to Healthcare IoT
        1. Decision science in health data
        2. Conclusion
      3. Smart healthcare - Connected Humans to Smart Humans
      4. Evolving from connected cars to smart cars
        1. Smart refuel assistant
        2. Predictive maintenance
        3. Autonomous transport
        4. Concluding thoughts
      5. Summary

Product information

  • Title: Smarter Decisions – The Intersection of Internet of Things and Decision Science
  • Author(s): Jojo Moolayil
  • Release date: July 2016
  • Publisher(s): Packt Publishing
  • ISBN: 9781785884191