TensorFlow (Python) implementation of DeepTCN model for multivariate time series forecasting.

Last update: Dec 19, 2022

Overview

DeepTCN TensorFlow

TensorFlow (Python) implementation of multivariate time series forecasting model introduced in Chen, Y., Kang, Y., Chen, Y., & Wang, Z. (2020). Probabilistic forecasting with temporal convolutional neural network. Neurocomputing, 399, 491-501. https://doi.org/10.1016/j.neucom.2020.03.011.

Dependencies

pandas==1.3.4
numpy==1.19.5
tensorflow==2.7.0
tensorflow_probability==0.14.1
plotly==5.3.1
kaleido==0.2.1

Usage

import numpy as np
from deep_tcn_tensorflow.model import DeepTCN

# Generate two time series
N = 1000
t = np.linspace(0, 1, N)
e = np.random.multivariate_normal(mean=[0, 0], cov=[[1, 0.25], [0.25, 1]], size=N)
a = 40 + 30 * t + 20 * np.cos(2 * np.pi * (10 * t - 0.5)) + e[:, 0]
b = 50 + 40 * t + 30 * np.sin(2 * np.pi * (20 * t - 0.5)) + e[:, 1]
y = np.hstack([a.reshape(- 1, 1), b.reshape(- 1, 1)])

# Fit the model
model = DeepTCN(
    y=y,
    x=None,
    forecast_period=100,
    lookback_period=200,
    quantiles=[0.01, 0.1, 0.5, 0.9, 0.99],
    filters=4,
    kernel_size=3,
    dilation_rates=[1, 2],
    loss='nonparametric'
)

model.fit(
    learning_rate=0.01,
    batch_size=64,
    epochs=200,
    verbose=1
)

# Plot the in-sample predictions
predictions = model.predict(index=900)
fig = model.plot_predictions()
fig.write_image('predictions.png', width=750, height=650)

# Plot the out of sample forecasts
forecasts = model.forecast()
fig = model.plot_forecasts()
fig.write_image('forecasts.png', width=750, height=650)

This repository contains the implementations related to the experiments of a set of publicly available datasets that are used in the time series forecasting research space.

TSForecasting This repository contains the implementations related to the experiments of a set of publicly available datasets that are used in the tim

80 Dec 30, 2022

The GitHub repository for the paper: “Time Series is a Special Sequence: Forecasting with Sample Convolution and Interaction“.

SCINet This is the original PyTorch implementation of the following work: Time Series is a Special Sequence: Forecasting with Sample Convolution and I

386 Jan 1, 2023

Time Series Forecasting with Temporal Fusion Transformer in Pytorch

TensorFlow (Python) implementation of DeepTCN model for multivariate time series forecasting.

Related tags

Overview

DeepTCN TensorFlow

Dependencies

Usage

You might also like...

This repository contains the implementations related to the experiments of a set of publicly available datasets that are used in the time series forecasting research space.

The GitHub repository for the paper: “Time Series is a Special Sequence: Forecasting with Sample Convolution and Interaction“.

Time Series Forecasting with Temporal Fusion Transformer in Pytorch

Event-forecasting - Event Forecasting Algorithms With Python

LONG-TERM SERIES FORECASTING WITH QUERYSELECTOR – EFFICIENT MODEL OF SPARSEATTENTION

Forecasting for knowable future events using Bayesian informative priors (forecasting with judgmental-adjustment).

Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting

A Tensorflow based library for Time Series Modelling with Gaussian Processes

Contra is a lightweight, production ready Tensorflow alternative for solving time series prediction challenges with AI

Owner

Flavia Giammarino

Code for the CIKM 2019 paper "DSANet: Dual Self-Attention Network for Multivariate Time Series Forecasting".

The source code and data of the paper "Instance-wise Graph-based Framework for Multivariate Time Series Forecasting".

Time-series-deep-learning - Developing Deep learning LSTM, BiLSTM models, and NeuralProphet for multi-step time-series forecasting of stock price.

Deep learning-based approach to discovering Granger causality networks in multivariate time series

USAD - UnSupervised Anomaly Detection on multivariate time series

Library for implementing reservoir computing models (echo state networks) for multivariate time series classification and clustering.

This repo contains the code required to train the multivariate time-series Transformer.

Implementation of the paper NAST: Non-Autoregressive Spatial-Temporal Transformer for Time Series Forecasting.

Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting

tsai is an open-source deep learning package built on top of Pytorch & fastai focused on state-of-the-art techniques for time series classification, regression and forecasting.