| Table of Contents |
|---|
Author(s)
| Name | Institution | Mail Address | Social Contacts |
|---|---|---|---|
| Lucio Anderlini | INFN Sezione di Firenze | Lucio.Anderlini@fi.infn.it | Hangouts: l.anderlini@gmail.com |
| Matteo Turisini | Università La Sapienza | Matteo.Turisini@roma1.infn.it | N/A |
How to Obtain Support
| Lucio.Anderlini@fi.infn.it | |
| Social | Hangouts: l.anderlini@gmail.com |
| Jira | N/A |
General Information
| ML/DL Technologies | Forward Neural Networks; Deployment in C |
|---|---|
| Science Fields | Generic |
| Difficulty | Medium |
| Language | English |
| Type | fully annotated |
Software and Tools
| Programming Language | Python, C |
|---|---|
| ML Toolset | Keras + Tensorflow |
| Additional libraries | scikit-learn |
| Suggested Environments | INFN-Cloud VM, bare Linux Node, Google CoLab |
Needed datasets
| Data Creator | scikit-learn community |
|---|---|
| Data Type | image |
| Data Size | < 1 GB |
| Data Source | scikit-learn package |
Short Description of the Use Case
In the life of a physicist, the time arrives when you get disgusted by the software dependencies that evaluating a simple neural network requires and you want the plain, pure and clean function to be evaluated in a whatever context.
...
The other reason is that the original data that we used to train this are not public.
How to execute it
Download and run the jupyter Static jupter notebook: https://github.com/landerlini/MLINFN-TutorialNotebooks/blob/master/TranscribeAnMLPinCLHCbMasterclassExplained.ipynb
Requirements
To run this exercise you will need a plain installation of python with tensorflow 1.x (should work with tensorflow 2.x, but it was not tested).
...