# Forecasting with LSTM + MLJ How to ? (A detailed guide or tutorial is needed)

**URL:** <https://discourse.julialang.org/t/forecasting-with-lstm-mlj-how-to-a-detailed-guide-or-tutorial-is-needed/94835>\
**Category:** Machine Learning\
**Tags:** mlj, lstm\
**Created:** [February 18, 2023, 1:43pm UTC](https://discourse.julialang.org/t/forecasting-with-lstm-mlj-how-to-a-detailed-guide-or-tutorial-is-needed/94835 "2023-02-18T13:43:14Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![huqedato](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/huqedato/32/46373_2.png) [@huqedato](https://discourse.julialang.org/u/huqedato)\
**Post date:** [February 18, 2023, 1:43pm UTC](https://discourse.julialang.org/t/forecasting-with-lstm-mlj-how-to-a-detailed-guide-or-tutorial-is-needed/94835/1 "2023-02-18T13:43:14Z")

</div>

For a future task/project (starting this late spring) we will be required to create and deploy a network for predictive analysis on a large time-series dataset with multiple complex features.  
The tools of work (set by my hierarchy/CTO) will be Julia env with MLJ/Flux libraries and LSTM model.

I’ve been assigned (as Product Architect and main developer) to lead this task and now I am at the prospective stage (understanding the problem, learning, studying, business needs, gathering information, designing the product, etc.).

My level of experience/skills with Julia is quite good but I’ve got absolutely no experience on ML/AI with Julia. I have general knowledge of ML, I had several past encounters with TensorFlow and specially with Google’s AutoML. My team consists of two senior developers (C++, Julia, Python) with no AI/ML experience, a data analyst and a theoretical AI/ML “_expert_” (a university-level scientist with no practical - application development skills).

Therefore I started to study/practice ML on Julia (plenty of documentation online). But I have the feeling that I’m going around in circles and the time is running out, I will be asked soon to deliver the draft with the architecture of the product including the business needs and resource requirements.

What I actually need is a precise step-by-step tutorial, preferably a full example on how to deploy a LSTM network on Julia from A to Z on a dummy dataset.

Thanks for you help.  
T.

---

<div class="post-metadata">

**Author:** ![bienpierre](https://avatars.discourse-cdn.com/v4/letter/b/8c91f0/32.png) [@bienpierre](https://discourse.julialang.org/u/bienpierre)\
**Post date:** [February 25, 2023, 2:35pm UTC](https://discourse.julialang.org/t/forecasting-with-lstm-mlj-how-to-a-detailed-guide-or-tutorial-is-needed/94835/2 "2023-02-25T14:35:08Z")

</div>

Hi,  
You can have a look on [model zoo](https://github.com/FluxML/model-zoo) for a step -by step examples with various recurrent neural networks.  
Regards