Time Series Analysis (Forecasting, Mining, Transformation, Clustering, Classification) + Python code

Python codes and datasets:

https://github.com/fanaee/TimeSeriesAnalysis

Covered topics:

1) Time Series Forecasting

-- Time Series Components
----- Level
----- Noise
----- Trend
----- Seasonality
----- Outlier
-- Stationarity
----- Seasonal-Trend Decomposition
----- Augmented Dickey-Fuller Test
----- Logarithmic/Square root/Box-Cox transform
----- Trend Removal via differencing
----- Fixing Non-constant Variance + Trend Removal
-- Statistical Models
----- Exponential Smoothing
----- Autoregressive Model (AR)
----- Moving Average Model (MA)
----- ARMA
----- ARIMA
----- SARIMA
-- State Space Models
----- Kalman Filter
----- Hidden Markov Model
-- Deep Learning Models
----- Feed-forward Neural Networks
----- One-dimensional Convolutional Neural Networks
----- Recurrent Neural Network
- Other Models: Facebook Prophet

2) Mining Time Series
Matrix Profile
Motifs
Discords

3) Time Series Transformation
-- Fourier Transform
-- Wavelets
-- Piecewise Aggregate Approximation (PAA)
-- Symbolic Aggregate Approximation (SAX)

4) Time Series Clustering/Classification

5) Preparation of Time Series Data

-- Missing Values
-- Noise Removal
-- Normalization
-- Resampling