Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Signal forecasting with a Sequence-to-Sequence (seq2seq) Recurrent Neural Network (RNN) model in TensorFlow - Guillaume Chevalier
| Date | Stars |
|---|---|
| 2026-07-24 | 1082 |
| 2026-07-25 | 1082 |
| 2026-07-28 | 1082 |
| 2026-07-30 | 1082 |
| 2026-07-31 | 1081 |
| 2026-08-06 | 1081 |
Today
— stars today
This week
-1 stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# [Sequence to Sequence (seq2seq) Recurrent Neural Network (RNN) for Time Series Forecasting](https://github.com/guillaume-chevalier/seq2seq-signal-prediction) ***Note: You can find here the accompanying [seq2seq RNN forecasting presentation's slides](https://drive.google.com/drive/folders/1U0xQMxVespjQilMhYW4mDxN02IwEW67I), as well as the Google Colab file for running the present notebook (if you're not already in Colab).*** This is a series of exercises that you can try to solve to learn how to code Encoder-Decoder Sequence to Sequence Recurrent Neural Networks (seq2seq RNNs). You can solve different simple toy signal prediction problems. Seq2seq architectures may also be used for other sophisticated purposes, such as for Natural Language Processing (NLP). In this project are given 4 exercises of gradually increasing difficulty. I take for granted that you have at least some knowledge of how RNN works and how can they be shaped into an encoder and a decoder seq2seq setup of the most simple form (without attention). To learn more about RNNs in TensorFlow, you may want to visit [this other RNN project](https://github.com/guillaume-chevalier/LSTM-Human-Activity-Recognition) which I have built for that. The current project is a series of example I have first built in French, but I haven't got the time to generate all the charts anew with proper English text. I have built this project at first for the practical part of the third hour of a [master class](https://webaquebec.org/classes-de-maitre/deep-learning-avec-tensorflow) conference I presented at the Web At Quebec (WAQ), originally in March 2017. ## How to use this ".ipynb" Python notebook ? I made available an ".py" Python version of this tutorial within the [repository](https://github.com/guillaume-chevalier/seq2seq-signal-prediction), but it's more convenient to run the code inside the notebook or within Google Colab. For running the notebook, you can run `jupyter-notebook` in the command-line to launch the web notebook IDE, and choose the `.ipynb` file. For Google Colab, if you want to run the code using GPU, make sure to do `Runtime > Change Runtime Type` and to select `GPU` for `Python 3`. ## Exercises Note that the dataset changes in function of the exercice. Most of the time, you will have to edit the neural networks' training parameter to succeed in doing the exercise, but at a certain point, changes in the architecture itself will be asked and required. The datasets used for this exercises are found in [`datasets.py`](https://github.com/guillaume-chevalier/seq2seq-signal-prediction/blob/master/datasets.py). ### Exercise 1 In theory, it is possible to create a perfect prediction of the signal for this exercise as it is deterministic. The neural network's parameters has been set to "somehow" acceptable values for a first training. You'll want to play with the hyperparameters until you reach predictions like those: <img src="https://github.com/guillaume-chevalier/seq2seq-signal-prediction/blob/master/images/E1.png?raw=true" /> Note: the neural network sees only what is to the left of the chart and is trained to predict what is at the right (predictions in yellow). We have 2 time series at once to predict, which are tied together. That means our neural network processes multidimensional data. A simple example would be to receive as an argument the past values of multiple stock market symbols in order to predict the future values of all those symbols with the neural network, which values are evolving together in time. That is what we will do in the exercise 4 with USD and EUR values of the BTC that we'll see both at once. ### Exercise 2 Here, rather than 2 signals in parallel to predict, we have only one, for simplicity. HOWEVER, this signal is a superposition of two sine waves of varying wavelenght and offset (and restricted to a particular min and max limit of wavelengts). In order to finish this exercise properly, you will need to edit the neural
Excerpt of 46,065 characters
Read on GitHub
56
Guillaume Chevalier · Canada
47
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:7b2b63c869f17b83, topic:tensorflow