Time series analysis
See:
- AI/Supervised Learning/Regression
- AI/Forecasting
- AI/Deep learning/RNNs
- “Sequence (time series) modelling” section in AI/Deep learning/CNNs
- “Deep learning for tabular data” section in AI/Deep learning/DL
- AI/Deep learning/Transformers
- AI/Causality
# Resources
- https://en.wikipedia.org/wiki/Time_series
- https://github.com/MaxBenChrist/awesome_time_series_in_python
- https://github.com/frutik/awesome-timeseries
- https://github.com/cuge1995/awesome-time-series
- http://machinelearningmastery.com/convert-time-series-supervised-learning-problem-python/
- Python to work with time series data
# Code
- See “Code” in AI/Forecasting
- #PAPER A systematic review of Python packages for time series analysis (Siebert 2021)
- #CODE tsai - State-of-the-art Deep Learning library for Time Series and Sequences
- #CODE Sktime
- #CODE Stumpy - STUMPY is a powerful and scalable Python library for modern time series analysis
- #CODE Orbit - A Python package for Bayesian forecasting with object-oriented design and probabilistic models under the hood
- #CODE TSflex
- #CODE Merlion - Merlion is a Python library for time series intelligence
- #CODE
Kats
- Kats, a kit to analyze time series data, a lightweight, easy-to-use, generalizable, and extendable framework to perform time series analysis, from understanding the key statistics and characteristics, detecting change points and anomalies, to forecasting future trends
- https://facebookresearch.github.io/Kats/
- https://engineering.fb.com/2021/06/21/open-source/kats/
- https://towardsdatascience.com/kats-a-generalizable-framework-to-analyze-time-series-data-in-python-3c8d21efe057
- #CODE Tsfresh - Time Series Feature extraction based on scalable hypothesis tests
- #CODE TSFEL
- #CODE Tslearn - A machine learning toolkit dedicated to time-series data
- #CODE
Pmdarima
- Pyramid bridges one more gap between R and Python by bringing R’s auto.arima to Python. Pyramid wraps statsmodels’ well-tested ARIMA and SARIMAX estimators.
- #CODE Tick
- #CODE Feature engine
- #CODE Featuretools - An open source python library for automated feature engineering
# Subtopics
# Time Series Forecasting
See AI/Forecasting
# Anomaly and Outlier Detection
See “For Time series” section in AI/Anomaly and Outlier Detection
# TS models
- Autoregressive
- Moving average
- Autoregressive moving average (ARMA)
- Autoregressive integrated moving average (ARIMA)
- Generalized additive model (GAM)
- #PAPER HyperTime: Implicit Neural Representation for Time Series (Fons 2022)
# TS classification
- UEA & UCR Time Series Classification Repository
- Dynamic time warping - DTW is one of the algorithms for measuring similarity between two temporal sequences, which may vary in speed
- https://medium.com/@hassanismailfawaz/deep-learning-for-time-series-classification-a-brief-overview-73b58767ed0f
- #PAPER ROCKET: Exceptionally fast and accurate time series classification using random convolutional kernels (Dempster 2019)
- #PAPER InceptionTime: Finding AlexNet for Time Series Classification (Ismail Fawaz 2019)
- #PAPER Deep learning for time series classification: a review (Ismail Fawaz 2019)
- #PAPER TS-CHIEF: A Scalable and Accurate Forest Algorithm for Time Series Classification (Shifaz 2020)
# Time-frequency analysis
# Fourier Analysis
- Fourier analysis
- Fast Fourier transform
- A fast Fourier transform (FFT) is an algorithm that computes the discrete Fourier transform (DFT) of a sequence, or its inverse (IDFT).
- Fourier Transform for Time Series
- Understanding FFTs and Windowing
- Fast Fourier transform
# Causality
See AI/Causality
# Explainability
See AI/XAI