Time Series Forecasting

LSTM for Time Series Forecasting

Univariate LSTM Models : one observation time-series data, predict the next value in the sequence
Multivariate LSTM Models : two or more observation time-series data, predict the next value in the sequence
- Multiple Input Series : two or more parallel input time series and an output time series that is dependent on the input time series
- Multiple Parallel Series : multiple parallel time series and a value must be predicted for each
Univariate Multi-Step LSTM Models : one observation time-series data, predict the multi step value in the sequence prediction.
Multivariate Multi-Step LSTM Models : two or more observation time-series data, predict the multi step value in the sequence prediction.
- Multiple Input Multi-Step Output.
- Multiple Parallel Input and Multi-Step Output.

Machine Learning for Multivariate Input

Statistical Method for Multivariate Input

Machine Learning for Univariate Input

Time Series Forecasting with the Long Short-Term Memory Network in Python : LSTM
rishikksh20/LSTM-Time-Series-Analysis: Using LSTM network for time series forecasting Dataset: Sunspot Zurich, Model: LSTM
sagarmk/Forecasting-on-Air-pollution-with-RNN-LSTM: Time Series Forecasting using LSTM in Keras. Dataset: Air Pollution, Model: LSTM
pushpendughosh/Stock-market-forecasting: Forecasting directional movements of stock prices for intraday trading using LSTM and random forest Dataset: Stock Market, Model: LSTM, RF
deshpandenu/Time-Series-Forecasting-of-Amazon-Stock-Prices-using-Neural-Networks-LSTM-and-GAN-: Project analyzes Amazon Stock data using Python. Feature Extraction is performed and ARIMA and Fourier series models are made. LSTM is used with multiple features to predict stock prices and then sentimental analysis is performed using news and reddit sentiments. GANs are used to predict stock data too where Amazon data is taken from an API as Generator and CNNs are used as discriminator. Dataset: Amazon Stock Model: LSTM with addition
demmojo/lstm-electric-load-forecast: Electric load forecast using Long-Short-Term-Memory (LSTM) recurrent neural network Dataset: Electric Consumption Model: LSTM
Yifeng-He/Electric-Power-Hourly-Load-Forecasting-using-Recurrent-Neural-Networks: This project aims to predict the hourly electricity load in Toronto based on the loads of previous 23 hours using LSTM recurrent neural network. Dataset: Electricity Consumption Model: LSTM
Yongyao/enso-forcasting: Improving the forecasting accuracy of ENSO through deep learning Dataset: ENSO El Nino, Model: LSTM
EsmeYi/time-series-forcasting: Using K-NN, SVM, Bayes, LSTM, and multi-variable LSTM models on time series forecasting Dataset: Sensor, Model: LSTM
CynthiaKoopman/Forecasting-Solar-Energy: Forecasting Solar Power: Analysis of using a LSTM Neural Network Dataset: Solar power, Model: LSTM
3springs/attentive-neural-processes: implementing "recurrent attentive neural processes" to forecast power usage (w. LSTM baseline, MCDropout) Dataset: English power consumption, Model: ANP-RNN "Recurrent Attentive Neural Process for Sequential Data", ANP: Attentive Neural Processes, NP: Neural Processes, LSTM
Housiadas/forecasting-energy-consumption-LSTM: Development of a machine learning application for IoT platform to predict electric energy consumption in smart building environment in real time., Dataset: Kaggle energy consuption, Model: LSTM, Seq2Seq

Statistical Method for Univariate Input

Time Series Forecasting—ARIMA, LSTM, Prophet with Python - by Caner Dabakoglu - Medium : LSTM, ARIMA, Prophet
pyaf/load_forecasting: Load forcasting on Delhi area electric power load using ARIMA, RNN, LSTM and GRU models Dataset: Electricity, Model: Feed forward Neural Network FFNN, Simple Moving Average SMA, Weighted Moving Average WMA, Simple Exponential Smoothing SES, Holts Winters HW, Autoregressive Integrated Moving Average ARIMA, Recurrent Neural Networks RNN, Long Short Term Memory cells LSTM, Gated Recurrent Unit cells GRU, Type: Univariate
jiegzhan/time-series-forecasting-rnn-tensorflow: Time series forecasting Dataset: Daily Temperature, Model: LSTM
zhangxu0307/time_series_forecasting_pytorch: time series forecasting using pytorch，including ANN,RNN,LSTM,GRU and TSR-RNN，experimental code Dataset: Pollution, Solar Energy, Traffic data etec. Model MLP,RNN,LSTM,GRU, ARIMA, SVR, RF and TSR-RNN
rakshita95/DeepLearning-time-series: LSTM for time series forecasting Dataset: ?? Model: ARIMA, VAR, LSTM
mborysiak/Time-Series-Forecasting-with-ARIMA-and-LSTM Dataset: Olypic, LeBron, Zika, Model: ARIMA dan LSTM
stxupengyu/load-forecasting-algorithms: 使用多种算法（线性回归、随机森林、支持向量机、BP神经网络、GRU、LSTM）进行电力系统负荷预测/电力预测。通过一个简单的例子。A variety of algorithms (linear regression, random forest, support vector machine, BP neural network, GRU, LSTM) are used for power system load forecasting / power forecasting. Dataset: Power usage, Model: linear regression, random forest, support vector machine, BP neural network, GRU, LSTM
Abhishekmamidi123/Time-Series-Forecasting: Rainfall analysis of Maharashtra - Season/Month wise forecasting. Different methods have been used. The main goal of this project is to increase the performance of forecasted results during rainy seasons. Dataset: precipitation, Model: ARIMA, LSTM, FNN(Feed forward Neural Networks), TLNN(Time lagged Neural Networks), SANN(Seasonal Artificial Neural Networks

Jupyter Notebook Examples

Univariate ARIMA

import statsmodels

Univariate LSTM

import keras

Multivariate VAR

(Note: VAR should only for Stationary process - Wikipedia)

Multivariate LSTM

Prophet and Kats from Facebook

Note on Multivariate and Univariate

Software

Gekko Timeseries and Modeling Software - Timeseries handling, and solving of large-scale economic models

Other Time Series

Time Series Forecasting with Regression and LSTM - Paperspace Blog
Kats - Kats One stop shop for time series analysis in Python
chlubba/catch22: catch-22: CAnonical Time-series CHaracteristics
blue-yonder/tsfresh: Automatic extraction of relevant features from time series:

Precipitation Forecasting

Nowcasting, Upsampling, Interpolation, Super resolution
hydrogo/rainnet: RainNet: a convolutional neural network for radar-based precipitation nowcasting
1706.03458 Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model
- sxjscience/HKO-7: Source code of paper "NIPS2017 Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model"

Deep Learning for Forecasting

top open source deep learning for time series forecasting frameworks.

Gluon This framework by Amazon remains one of the top DL based time series forecasting frameworks on GitHub. However, there are some down sides including lock-in to MXNet (a rather obscure architecture). The repository also doesn’t seem to be quick at adding new research.
Flow Forecast This is an upcoming PyTorch based deep learning for time series forecasting framework. The repository features a lot of recent models out of research conferences along with an easy to use deployment API. The repository is one of the few repos to have new models, coverage tests, and interpretability metrics.
sktime dl This is another time series forecasting repository. Unfortunately it looks like particularly recent activity has diminished on it.
PyTorch-TS Another framework, written in PyTorch, this repository focuses more on probabilistic models. The repository isn’t that active (last commit was in November).

eBook Forecasting

Timeseries Forecasting

Timeseries Forecasting Book

Timeseries Forecasting Reading

Timeseries RNN

Spice.ai

Timeseries Forecasting

Time-series Forecasting

VAR

time Series

LSTM

LSTM on Pytorch

Time Series Toolbox

Prophet Prophet | Forecasting at scale.
Forecasting Models for Tidy Time Series • fable in R
unit8co/darts: A python library for easy manipulation and forecasting of time series. in Python
CRAN - Package forecast in R robjhyndman/forecast: forecast package for R
XGboost, LGBM, pmdarima, stanpy (for bayesian modelling)
Prophet - seems to be the current 'standard' choice
ARIMA - Classical choice
Exponential Moving Average - dead simple to implement, works well for stuff that's a time series but not very seasonal
Kalman/Statespace model - used by Splunk's predict[1] command (pretty sure I always used LLP5)
Prophet, statsmodels, tf.keras for RNNs.
tensorflow probability's time series package.
PyTorch for recurrent nets
tsfresh—tsfresh 0.18.1.dev39+g611e04f documentation
MiniRocket—sktime documentation
State of the art is 1D convnets, bleeding edge is transformers.
pycaret timeseries
lgbm light gbm
cvxpy
TensorFlow's LSTMCell
LSTMs have been going the way of the dinosaurs since 2018. If you really need a complex neural network (over 1D convolution approaches), transformers are the current SOTA. Demand forecasting with the Temporal Fusion Transformer—pytorch-forecasting documentation
sktime
bssts
statsmodels
https://github.com/fraunhoferportugal/tsfel

Books

Forecasting: Principles and Practice (2nd ed)

Forecasting Comparison

statsforecast/experiments/m3 at main · Nixtla/statsforecast # Statistical vs Deep Learning forecasting methods
Nixtla/statsforecast: Lightning ⚡️ fast forecasting with statistical and econometric models.

PreviousSolar Energy NextVirtual Reality and Augmented Reality

Last updated 2 years ago

Was this helpful?