책 《 실전 시계열 분석 Practical Time Series Analysis》데이터 자료 정리

2023-02-16 1차 수정

소스코드

https://deep-diver.github.io/practical-time-series-analysis-korean/

책 공식 저장소

https://github.com/PracticalTimeSeriesAnalysis/BookRepo

 

1장

life table (https://perma.cc/HU6A-9W22) 

천문학 이미지 관찰  (https://perma.cc/2TNK-2TFW)

 

시계열 분석과 예측의 역사

• ⌜A brief history of forecasing competitions⌟  (https://perma.cc/32LJ-RFJW) - 시계열을 사용한 예측 대회가 어떻게 컴퓨터와 비슷한 속도로 발전했는지의 예
• ⌜Revisiting Francis Galton’s Forecasting Competition⌟ (https://perma.cc/FJ6V-8HUY) - 마을 장날 산 채로 도살된 소의 무게를 예측
• ⌜"On a method of investigating periodicities disturbed series, with special reference to Wolfer's sunspot numbers." Philosophical Transactions of the Royal Society of London. Series A, Containing Papers of a Mathematical or Physical Character⌟ (https://perma.cc/D6SL-7UZS) - 처음으로 실제 데이터에 ARIMA 적용. 주기적인 현상의 추정 분석법에서 주기성의 가정을 제거하는 방법 설명
• ⌜The Combination of Forecasts⌟ (https://perma.cc/9AEE-QZ2J) - 시계열 분석을 위한 앙상블 기법 사용

⌜25 years of time series forecasting⌟ (https://perma.cc/84RG-58BU) - 20세기 시계열 예측 연구를 통계적으로 요약

 

특정 도메인의 시계열 역사와 논의

• ⌜Weather Forecasting Through the Ages⌟ (https://perma.cc/8GK5-JAVT)
• ⌜Early Meteorological Data from London and Paris:Extending the North Atlantic Oscillation Series⌟ (https://perma.cc/NJ33-WVXH)
• ⌜A brief history of medicine and statistics⌟ (https://perma.cc/WKU3-9SUX)
• ⌜Random time series in Astronomy⌟ (https://perma.cc/J3VS-6JYB)

 

2장

UEA, UCR 시계열 분류 저장소 (https://perma.cc/56Q5-YPNT)

요가 동작 분류 작업 (https://perma.cc/U6MU-2SCZ)

와인 데이터셋 (https://perma.cc/Y34R-UGMD)

CRAN 저장소 (https://perma.cc/2694-D79K)

Pandas 개요(https://perma.cc/7R9B-2YPS)

 

 - R

누락 데이터 대치법 시험 데이터 (https://data.bls.gov/timeseries/LNS14000000)

R의 data.table 공식 문서 (https://perma.cc/3HEB-NE6A)

R 시계열 패키지 XTS (https://perma.cc/83E9-4N79)

 

누락된 데이터

• ⌜Comparison of different Methods for Univariate Time Series Imputation in R⌟ (https://perma.cc/M4LJ-2DFB)
• ⌜What to Do about Missing Values in Time-Series Cross-Section Data⌟ (https://perma.cc/8ZLG-SMSX)
• ⌜Time series - Notes on irregular time series and missing values⌟ (https://perma.cc/8LHP-92FP)
• ⌜The Problem with Time & Timezones - Computerphile⌟ (https://www.youtube.com/watch?v=-5wpm-gesOY)
• ⌜Wikipedia, "Time zone"⌟ (https://perma.cc/J6PB-232C)
• ⌜GPS glitch threatens thousands of scientific instruments⌟ (https://www.nature.com/articles/d41586-019-01048-2)

 

평활과 계절성

• ⌜Chapter 7 Exponential smoothing⌟ (https://perma.cc/UX4K-2V5N)
• ⌜The correct way to start an Exponential Moving Average (EMA)⌟ (https://perma.cc/ZPJ4-DJJK)
• ⌜Time Series Using Exponential Smoothing Cells⌟ (https://perma.cc/2JRX-K2JZ)

 

함수적 데이터 분석

• ⌜Review of Functional Data Analsis⌟ (https://perma.cc/3DNT-J9EZ)
• ⌜Applications of functional data analysis: A systematic review⌟ (https://perma.cc/VGK5-ZEUX)

 

3장

⌜Why do we Sometimes get Nonsense-Correlations between Time-Series?--A Study in Sampling and the Nature of Time-Series⌟ (https://www.jstor.org/stable/2341482)

⌜spurious correlations⌟ (https://perma.cc/6UYH-FPBX)

⌜Fiftywords dataset⌟ (https://perma.cc/Y982-9FPS)

 

허위상관

• ⌜Ai Deng "A Primer on Spurious Statistical Significance in Time Series Regressions"⌟ (https://perma.cc/9CQR-RWHC)
• ⌜Tyler Vigen, Spurious Correlations⌟ (https://perma.cc/YY6R-SKWA)
• ⌜Spurious Regression Under Broken‐Trend Stationarity⌟ (https://perma.cc/V993-SF4F)
• ⌜Spurious regressions in econometrics⌟ (https://perma.cc/M8TE-AL6U)

탐색적 데이터 분석

• ⌜An exploratory data analysis of the temperature fluctuations in a spreading fire⌟ (https://perma.cc/QB3D-APKM)
• ⌜Time Series Regression and Exploratory Data Analysis⌟ (https://perma.cc/UC5B-TPVS)

 

더 많은 시각화 자료

• ⌜The TimeViz Browser A Visual Survey of Visualization Techniques for Time-Oriented Data⌟ (https://perma.cc/94ND-6ZA5)
• ⌜Displaying time series, spatial and space-time data with R (1st Edition)⌟ (https://perma.cc/R69Y-5JPL)
• ⌜5 Ways to Do 2D Histograms in R⌟ (https://perma.cc/ZCX9-FQQY)

 

여러 가지 추세

• ⌜Consideration of Trends in Time Series⌟ (https://perma.cc/WF2H-TVTL)

 

4장 시계열 데이터의 시뮬레이션

• ⌜Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs⌟ (https://perma.cc/Q69W-L44Z) - 윤리적이고 합법적이며 사생활보호적인 의료 데이터셋을 생성하는 데 딥러닝 시뮬레이션이 사용되는 예
• ⌜Forecasting ocean waves: Comparing a physics-based model with statistical models⌟ (https://perma.cc/89DJ-ZENZ) - 물리적, 통계적 시스템 모델링에 대해 설명
• ⌜Bootstrap Methods for Time Series⌟ (https://perma.cc/6CQA-EG2E) - 시간의 의존성을 지닌 시계열 데이터를 통계적으로 시뮬레이션하는 것의 어려움

 

5장 시간 데이터 저장

시계열 데이터베이스 기술

• ⌜Thoughts on Time-series Databases⌟ (https://perma.cc/8GDC-6CTX)
• ⌜List of Time Series Databases⌟ (https://perma.cc/9SCQ-9G57)
• ⌜Percona Blog Poll: What Database Engine Are You Using to Store Time Series Data?⌟ (https://perma.cc/5PXF-BF7L)
• ⌜The State of the Time Series Database Market⌟ (https://perma.cc/WLA7-ABRU)
• ⌜"Prometheus Documentation: COMPARISON TO ALTERNATIVES"⌟ (https://perma.cc/M83E-NBHQ)

 

일반적인 데이터베이스 기술에 적응하는 것

• ⌜Storing Time Series in PostgreSQL Efficiently⌟ (https://perma.cc/QP2D-YBTS)
• ⌜Using Redis as a Time Series Database: Why and How⌟ (https://perma.cc/RDZ2-YM22)
• ⌜Time-series data: Why (and how) to use a relational database instead of NoSQL⌟ (https://perma.cc/A6CU-6XTZ)
• ⌜Combining multiple tables with valid from/to date ranges into a single dimension⌟ (https://perma.cc/B8CT-BCEK)

 

6장 시계열의 통계 모델

⌜ARIMA models for time series forecasting⌟ (https://perma.cc/P9BK-764B)

⌜하인드먼 방법⌟ (https://perma.cc/G4EG-6SMP)

 

고전 텍스트

• ⌜Forecasting: Principles and Practice⌟ (https://perma.cc/9JNK-K6US)
• ⌜Ruey Tsay, Analysis of Financial Time Series(2001)⌟ ()
• ⌜Robert H.Shumway, TIme Series Analysis and Its Applications(2017)⌟ ()

경험적 지침

• ⌜Summary of rules for identifying ARIMA models⌟ (https://perma.cc/37BY-9RAZ)
• ⌜Box-Jenkins Models⌟ (https://perma.cc/3XSC-Y7AG)
• ⌜The ARIMAX model muddle⌟ (https://perma.cc/4W44-RQZB)
• ⌜Regularization for ARIMA models⌟ (https://perma.cc/G8NQ-RCCU)

7장 시계열의 상태공간 모델

 

⌜An Explanation of the Kalman Filter⌟ (https://perma.cc/27RK-YQ52)

HMM의 실제 사용 사례

• ⌜Market Regime Detection using Hidden Markov Models in QSTrader⌟ (https://perma.cc/JRT2-ZDVJ)
• ⌜Hidden Markov Models and their Applications in Biological Sequence Analysis⌟ (https://perma.cc/4V4A-53TZ)
• ⌜ECG signal analysis through hidden Markov models⌟ (https://perma.cc/G37Y-XBQH)

 

칼만 필터 및 선형 가우스 상태 공간 모델

• ⌜An Introduction to the Kalman Filter⌟ (https://perma.cc/ZCU8-MXEF)
• ⌜A New Approach to Linear Filtering and Prediction Problems1⌟ (https://perma.cc/GNC4-YLEC)
• ⌜Kalman-and-Bayesian-Filters-in-Python⌟ (https://perma.cc/CMU5-Y94A)
• ⌜State-space models’ dirty little secrets: even simple linear Gaussian models can have estimation problems⌟ (https://perma.cc/9D8V-Z7KJ)

은닉 마르코프 모형

• ⌜Hidden Markov Models, Lecture Notes,Andrew W. Moore Professor School of Computer Science Carnegie Mellon University⌟ (https://perma.cc/K3HP-28T8)
• ⌜CS 188: Artificial Intelligence:Hidden Markov Models⌟ (https://perma.cc/V7U4-WPUA)
• ⌜What is the difference between the forward-backward and Viterbi algorithms?⌟ (https://perma.cc/QNZ5-U3CN)

 

베이즈 구조적 시계열

• ⌜Bayesian Time Series Analysis⌟ (https://perma.cc/578D-XCVH)
• ⌜Predicting the Present with Bayesian Structural Time Series⌟ (https://perma.cc/4EJX-6WGA)
• ⌜Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors⌟ (https://perma.cc/BRP8-Y33X)

8장 시계열 특징의 생성 및 선택

catch22: CAnonical Time-series CHaracteristics (https://perma.cc/57AG-V8NP)

 

특징 기반 시계열 분석

• ⌜Feature-based time-series analysis⌟ (https://perma.cc/6LZ6-S3NC)

 

특징 선택

• ⌜Time Series FeatuRe Extraction on basis of Scalable Hypothesis tests (tsfresh – A Python package)⌟ (https://oreil.ly/YDBM8)
• ⌜Distributed and parallel time series feature extraction for industrial big data applications ⌟ (https://arxiv.org/pdf/1610.07717.pdf)

 

특정 도메인 특징

• ⌜Technical analysis⌟ (https://perma.cc/8533-XFSZ)
• ⌜Methods of EEG Signal Features Extraction Using Linear Analysis in Frequency and Time-Frequency Domains⌟ (https://perma.cc/465U-QT53)
• ⌜From FATS to feets: Further improvements to an astronomical feature extraction tool based on machine learning⌟ (https://perma.cc/8ZEM-Y892)
• ⌜Supplementary Materials for Scalable and accurate deep learning for electronic health records⌟ (https://perma.cc/2LKM-326C)
• Embeddings( https://perma.cc/3KAZ-9A3Y )

 

주기도

• ⌜The Periodogram⌟ (https://perma.cc/5DRZ-VPR9)
• ⌜Understanding the Lomb-Scargle Periodogram⌟ (https://arxiv.org/pdf/1703.09824.pdf)

9장 시계열을 위한 머신러닝

⌜Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: Dependence on recording region and brain state⌟ (https://perma.cc/YZD5-CTJF)

⌜Decision Trees in Machine Learning⌟ (https://perma.cc/G9AA-ANEN)

⌜An Introduction to Clustering and different methods of clustering⌟ (https://perma.cc/36EX-3QJU)

⌜Choosing a clustering method⌟ (https://perma.cc/MHL9-2Y8A)

⌜Time series distance metric⌟ (https://perma.cc/389W-68AH)

 

시계열 거리 및 유사성 측정

• ⌜Dynamic Time Warping⌟ (https://perma.cc/R24Q-UR84)
• ⌜Computing the Fréchet distance between two polygonal curves⌟ (https://perma.cc/5QER-Z89V)
• ⌜Pjotr Roelofsen,"Time series clustering"⌟ (https://perma.cc/K8HJ-7FFE)
• ⌜An Empirical Evaluation of Similarity Measures for Time Series Classification⌟ (https://perma.cc/G2J4-TNMX)

시계열을 위한 머신러닝

• ⌜Introduction to Time Series Mining⌟ (https://perma.cc/ZM9L-NW7J)
• ⌜The M4 Competition: Results, findings, conclusion and way forward⌟ (https://perma.cc/42HZ-YVUU)

 

10장 시계열을 위한 딥러닝

• ⌜Gluon Neural Network Layers⌟ (https://perma.cc/8PQW-4NKY)
• ⌜FEED-FORWARD NETWORKS WITH ATTENTION CAN SOLVE SOME LONG-TERM MEMORY PROBLEMS⌟ (https://arxiv.org/pdf/1512.08756.pdf)
• ⌜How to calculate a Gaussian kernel matrix efficiently in numpy?⌟ (https://perma.cc/8U8Y-RBYW)

역사문헌

• ⌜LONG SHORT-TERM MEMORY⌟ (https://perma.cc/AHR3-FU5H)
• ⌜Forecasting With Artificial Neural Networks: The State of the Art⌟ (https://perma.cc/Z32G-4ZQ3)

RNN

• ⌜Chapter 15 – Processing Sequences Using RNNs and CNNs, handson-ml2⌟ (https://perma.cc/D3UG-59SX)
• ⌜DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks⌟ (https://perma.cc/MT7N-A2L6)
• ⌜Deep and Confident Prediction for Time Series at Uber⌟ (https://perma.cc/PV8R-PHV4)
• ⌜Recurrent Neural Networks for Multivariate Time Series with Missing Values⌟ (https://perma.cc/4YM4-SFNX)

CNN

⌜WaveNet: A Generative Model for Raw Audio⌟ (https://perma.cc/G37Y-WFCM)

딥러닝의 응용

• ⌜Automated Website Fingerprinting through Deep Learning⌟ (https://perma.cc/YR2G-UJUW)
• ⌜Web Traffic Time Series Forecasting⌟ (https://perma.cc/UUR4-VNEU)

 

11장 오차 측정

• ⌜A Note on the Validity of Cross-Validation for Evaluating Autoregressive Time Series Prediction⌟ (https://perma.cc/YS3J-6DMD)
• ⌜Bootstrapping of Time Series Description⌟ (https://perma.cc/MQ77-U5HL)
• ⌜The Jackknife and the Bootstrap for General Stationary Observations⌟ (https://perma.cc/5B2T-XPBC)
• ⌜Structural Change in (Economic) Time Series⌟ (https://perma.cc/7U8N-T4RC)
• ⌜Predictive regressions⌟ (https://perma.cc/YD7U-RXBM)

 

⌜⌟ ()

⌜⌟ ()

 

 

12장 시계열 모델의 학습과 배포에 대한 성능 고려 사항

균등한 성능을 보이는 모델

• ⌜The Great Time Series Classification Bake Off: An Experimental Evaluation of Recently Proposed Algorithms. Extended Version⌟ (https://perma.cc/T76B-M635)

간단한 모델 만들기

• ⌜Sequence-Level Knowledge Distillation⌟ (https://perma.cc/V4U6-EJNU)

 

13장 헬스케어 애플리케이션

• ⌜Advances in nowcasting influenza-like illness rates using search query logs⌟ (https://perma.cc/NQ6B-RUXF)
• ⌜Modeling the Past, Present, and Future of Influenza⌟ (https://perma.cc/96CZ-5SX2)
• ⌜Dynamic Regression⌟ (https://perma.cc/5TPY-PYZS) <- 파일 없음
• ⌜Dynamic Regression⌟ (https://robjhyndman.com/nyc2018/3-1-Dynamic-Regression.pdf)

 

14장 금융 애플리케이션

⌜야후 파이낸스 일간 기록 데이터⌟ (https://perma.cc/RQ6D-U4JX)

⌜A Clockwork RNN⌟ (https://perma.cc/9C62-7GFK)

• ⌜Multi-Task Learning for Stock Selection⌟ (https://perma.cc/GR7A-5PQ5)
• ⌜Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks⌟ (https://perma.cc/GJZ5-4V6Z)
• ⌜"Is anyone making money by using deep learning in trading?", Quora⌟ (https://perma.cc/Z8C9-V8FX)

 

 

15장 정부를 위한 시계열

• ⌜Computing Extremely Accurate Quantiles Using t-Digests⌟ (https://perma.cc/Z2A6-H76H)
• ⌜Quantile Curve Estimation and Visualization for Nonstationary Time Series⌟ (https://perma.cc/Z7T5-PSCB)
• ⌜Online Machine Learning in Big Data Streams⌟ (https://perma.cc/9TTY-VQL3)
• ⌜CSE 291: Unsupervised learning Spring 2008 Lecture 6 — Online and streaming algorithms for clustering⌟ (https://perma.cc/V3XL-GPK2)
• ⌜A Multi-Horizon Quantile Recurrent Forecaster⌟ (https://perma.cc/22AE-N7F3)

16장 시계열 패키지

• ⌜Is it possible to automate time series forecasting?⌟ (https://perma.cc/E3C4-RL4L)
• ⌜CausalImpact, google's open source pachage for causal inference⌟ (https://perma.cc/Y72Z-2SFD)
• ⌜Large-Scale Parallel Statistical Forecasting Computations in R⌟ (https://perma.cc/25D2-RVVA)
• ⌜Amazon Forecast – Time Series Forecasting Made Easy⌟ (https://perma.cc/Y2PE-EUDV)
• ⌜DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks⌟ (https://perma.cc/DNF9-LJKC)