An Introduction to IoT Analytics covers techniques that can be used to analyze data from IoT sensors and also addresses questions regarding the performance of an IoT system. It strikes a balance between practice and theory so that one can learn how to apply these tools in practice with a good understanding of their inner workings. It is an introductory book for readers that have no familiarity with these techniques.
The techniques presented in the book come from the areas of Machine Learning, Statistics, and Operations Research. Machine Learning techniques are described that can be used to analyze IoT data generated from sensors for clustering, classification, and regression. The statistical techniques described can be used to carry out regression and forecasting of IoT sensor data, and dimensionality reduction of data sets. Operations Research is concerned with the performance of an IoT system by constructing a model of a system under study, and then carry out what-if analysis. The book also describes simulation techniques.
- IoT analytics is not just Machine Learning but it also involves other tools, such as, forecasting and simulation techniques.
- Many diagrams and examples are given throughout the book to better explain the material presented.
- At the end of each chapter, there is a project designed to help the reader to better understand the techniques described in the chapter.
- The material is this book has been class tested over several semesters.
- Contains practice exercises, with solutions provided online at www.routledge.com/9780367686314
Table of Contents
1. Introduction. 1.1. The Internet of Things (IoT) . 1.2. IoT Application Domains. 1.3. IoT Reference Model. 1.4. Performance Evaluation and Modeling of IoT Systems. 1.5. Machine Learning and Statistical Techniques for IoT. 1.6. Overview of the Book. 2. Review of Probability Theory. 2.1. Random Variables. 2.2. Discrete Random Variables. 2.3. Continuous Random Variables 2.4. The Joint Probability Distribution. 3. Simulation Techniques. 3.1. Introduction. 3.2. The Discrete-event Simulation Technique. 3.3. Generating Random Numbers. 3.4. Simulation Designs. 3.5. Estimation Techniques. 3.6. Validation of a Simulation Model. 3.7. Simulation Languages. 4. Hypothesis Testing. 4.1. Statistical Hypothesis Testing for a Mean. 4.2. Analysis of Variance (ANOVA). 5. Multivariable Linear Regression. 5.1. Simple Linear Regression. 5.2. Multivariable Linear Regression. 5.3. An Example. 5.4. Polynomial Regression. 5.5. Confidence and Prediction Intervals. 5.6. Ridge, Lasso, and Elastic Net Regression. 6. Time Series Forecasting. 6.1. A Stationary Time Series. 6.2. Moving Average or Smoothing Models. 6.3. The Moving Average MA(q) Model. 6.4. The Autoregressive Model. 6.5. The Non-seasonal ARIMA (p,d,q) Model. 6.6. Decomposition Models. 6.7. Forecast Accuracy. 6.8. Prediction Intervals. 6.9. Vector Autoregression. 7. Dimensionality Reduction. 7.1. A Review of Eigenvalues and Eigenvectors. 7.2. Principal Component Analysis (PCA). 7.3. Linear and Multiple Discriminant Analysis. 8. Clustering Techniques. 8.1. Distance Metrics. 8.2. Hierarchical Clustering. 8.3. The k-means Algorithm. 8.4. The Fuzzy c-means Algorithm. 8.5. The Gaussian Mixture Decomposition. 8.6. The DBSCAN Algorithm. 9. Classification Techniques. 9.1. The k-nearest Neighbor (k-NN) Method. 9.2. The Naïve Bayes Classifier. 9.3. Decision Trees. 9.4. Logistic Regression. 10.: Artificial Neural Networks. 10.1. The Feedforward Artificial Neural Network. 10.2. Other Artificial Neural Networks . 10.3. Activation Functions. 10.4. Calculation of the Output Value. 10.5. Selecting the Number of Layers and Nodes . 10.6. The Backpropagation Algorithm. 10.7. Stochastic, Batch, Mini-batch Gradient Descent Methods. 10.8. Feature Normalization. 10.9. Overfitting. 10.10. Selecting the Hyper-parameters. 11. Support Vector Machines. 11.1. Some Basic Concepts. 11.2. The SVM Algorithm: Linearly Separable Data. 11.3. Soft -margin SVM (C-SVM). 11.4. The SVM Algorithm: Non-linearly Separable Data. 11.5. Other SVM methods. 11.6. Multiple Classes. 11.7. Selecting the Best Values for C and γ. 11.8. ε-Support Vector Regression (ε-SVR). 12. Hidden Markov Models. 12.1 Markov Chains. 12.2. Hidden Markov Models – An Example. 12.3. The Three Basic HMM Problems. 12.4. Mathematical Notation. 12.5. Solution to Problem 1. 12.6. Solution to Problem 2. 12.7. Solution to Problem 3. 12.8. Selection of the number of states N. 12.9. Forecasting OT+t. 12.10. Continuous Observation Probability Distributions. 12.11. Autoregressive HMMs. Appendix A: Some Basic Concepts of Queueing Theory. Appendix B: Maximum Likelihood Estimation (MLE).
Harry G. Perros is a Professor of Computer Science at NC State University, an Alumni Distinguished Graduate Professor, and an IEEE Fellow. He has published extensively in the area of performance modelling of computer and communication systems, and in his free time he likes to go sailing and play the bouzouki.