CHAPTER 1 Overview and History of Time Series

1.1 Time series history in various application fields
1.2 A leap forward in time series analysis
1.3 Origins of Statistical Time Series Analysis
1.4 The Origins of Machine Learning Time Series Analysis
1.5 Supplementary Materials

CHAPTER 2 Discovering and Handling Time Series Data

2.1 Where to find time series data
2.2 Improving Time Series Datasets in Table Sets
2.3 Problems with Timestamps
2.4 Data Cleaning
2.5 Seasonal data
2.6 Time zone
2.7 Prevention of pre-observation
2.8 Supplementary Materials

CHAPTER 3 Exploratory Data Analysis of Time Series

3.1 Familiar Method
3.2 Exploration methods specialized for time series
3.3 Useful Visualizations
3.4 Supplementary Materials

CHAPTER 4 SIMULATION OF TIME SERIES DATA

4.1 Special Features of Time Series Simulation
4.2 Simulation through code
4.3 Final advice on simulation
4.4 Supplementary Materials

CHAPTER 5 Time Data Storage

5.1 Requirements Definition
5.2 Database Solutions
5.3 File Solutions
5.4 Supplementary Materials

CHAPTER 6 Statistical Models for Time Series

6.1 Why not use linear regression?
6.2 Statistical models developed for time series
6.3 Advantages and Disadvantages of Time Series Statistical Models
6.4 Supplementary Materials

CHAPTER 7 State-Space Models of Time Series

7.1 Advantages and Disadvantages of State-Space Models
7.2 Kalman filter
7.3 Hidden Markov Models
7.4 Bayesian Structured Time Series
7.5 Supplementary Materials

CHAPTER 8 Generation and Selection of Time Series Features

8.1 Examples for Beginners
8.2 Considerations when calculating features
8.3 List of places that inspired the discovery of features
8.4 How to select only some of the generated features
8.5 In conclusion
8.6 Supplementary Materials

CHAPTER 9 Machine Learning for Time Series

9.1 Time series classification
9.2 Clustering
9.3 Supplementary Materials

CHAPTER 10 DEEP LEARNING FOR TIME SERIES

10.1 Deep Learning Concepts
10.2 Neural Network Programming
10.3 Creating a Learning Pipeline
10.4 Forward propagation network
10.5 Convolutional Neural Networks
10.6 Recurrent Neural Networks
10.7 Composite Structures
10.8 In conclusion
10.9 Supplementary Materials

CHAPTER 11 Error Measurement

11.1 Basic Methods for Testing Predictions
11.2 A good time to predict
11.3 Estimating model uncertainty through simulation
11.4 Predicting several steps ahead
11.5 Precautions when verifying models
11.6 Supplementary Materials

CHAPTER 12 Performance Considerations for Training and Deploying Time Series Models

12.1 Working with Tools Built for Common Cases
12.2 Advantages and Disadvantages of Data Storage Formats
12.3 Modifying the analysis to accommodate performance considerations
12.4 Supplementary Materials

CHAPTER 13 Healthcare Applications

13.1 Flu Forecast
13.2 Blood Sugar Level Prediction
13.3 Supplementary Materials

CHAPTER 14 Financial Applications

14.1 Acquisition and exploration of financial data
14.2 Preprocessing Financial Data for Deep Learning
14.3 Building and Training RNNs
14.4 Supplementary Materials

CHAPTER 15 TIME SERIES FOR GOVERNMENT

15.1 Obtaining government data
15.2 Exploring Large-Scale Time Series Data
15.3 Real-time statistical analysis of time series data
15.4 Supplementary Materials

CHAPTER 16 Time Series Package

16.1 Large-scale forecasting
16.2 and above detection
16.3 Other time series packages
16.4 Supplementary Materials

CHAPTER 17: The Future of Time Series Forecasting

17.1 Service-type forecasting
17.2 Improving Probabilistic Probability with Deep Learning
17.3 Machine learning methods are becoming more important than statistical methods.
17.4 The rise of methods combining machine learning and statistics.
17.5 More predictions permeate everyday life

A Practical Guide to Time Series Data Analysis from A to Z, Covering Amazon's #1 Data Warehouse

Time series analysis is used to predict and prepare for the future in places closely related to our daily lives, such as the Korea Meteorological Administration, financial institutions, and government agencies.
Time series data is becoming increasingly important due to the massive data production driven by the Internet of Things, the digital transformation of healthcare, and the rise of smart cities, and its impact will expand across all industries.

This book provides a comprehensive, practical overview of the entire time series data and modeling pipeline—acquisition, cleaning, simulation, storage, and modeling—for accurate time series analysis and forecasting, with R and Python code provided.
The first half introduces the concepts that are fundamental to understanding the entire process of time series forecasting.
It covers exploration, collection, and organization of time series data, as well as ARIMA and SARIMA models.
In the second half, you'll learn how to apply time series techniques to research cases in healthcare, finance, and government data using MXNet and TensorFlow, and you'll also learn about various examples that draw on the author's extensive experience.


Finally, we've included links to tutorials on the topics and essential techniques covered in each chapter.
This one book will prepare you to analyze and forecast time series data in the real world.
We hope this practical guide will help you improve the accuracy of your time series forecasts.

Main contents

Exploring and organizing time series data
Performing exploratory time series data analysis
Time data storage
Time series data simulation
Creating and selecting time series features
Measurement error
Time series forecasting and classification using machine learning and deep learning
Accuracy and Performance Evaluation

Example code

github.com/deep-diver/practical-time-series-analysis-korean