Enhancements to the ML models

[mvarlese]

Project Title: Phoebe - help increase data points density and data cleaning

Description: Phoeβe (/ˈfiːbi/) wants to add basic artificial intelligence capabilities to the Linux OS. Phoeβe uses system telemetry as the input to its brain and produces a big set of settings which get applied to the running system. The decision made by the brain is continuously reevaluated (considering the grace_period setting) to offer eventually the best possible setup.

Deliverable: Enhancements to the ML models, help increasing data points density and data cleaning

Mentor: Marco Varlese marco.varlese@suse.com

Skills: Must have: C, Python, Meson, Linux - Nice-to-have: artificial-intelligence, machine-learning

Skill Level: Medium

Get started: Please, clone the code repository [1], build it with different flags set and run the code in both training and inference mode.

[1] https://github.com/SUSE/phoebe

[guptadhaval18]

@mvarlese Hi I am interested in working on this issue. Could you please explain how the Enhancements to the ML models is to be done if the model used is x*w+b as specified in readme. Also what is data point density.
Also please provide link to the data to be cleaned.

[mvarlese]

@guptadhaval18 thanks for your interest.

The idea is to identify a model which could be better suited for few reasons:

• Keep providing a convergence to a meaningful result
• Have the model created in one of the ML frameworks, eg. Tensorflow or Pytorch, so that it could be easily updated and tested

For the new model we would be starting from the ones available in the aforementioned ML frameworks and see which once better suit us. Basically, let's try to use some existing algorithms for now.

We are currently lacking the amount of data which is really required for a thorough ML model hence we are setting up a lab where we can artificially create use case scenarios. Eventually, that data will be used to train the model for a real production use case. All that, to answer your last question about data density: we are still missing that information currently.

[guptadhaval18]

@mvarlese Thanks for quick response. Will be going through the repository. Will ask further questions if there is any doubt.

[mukul-mehta]

@mvarlese Hello, I'm interested in working on this project. To begin with, I'll build the project locally and try running it. Are there any other tasks I could take up to get more familiar with the project?

Also, is there a communication channel for discussions regarding this project?

[mvarlese]

@mukul-mehta thanks for expressing your interest. Yes, please, let's start from the activity you already depicted and possibly try to develop a basic plugin with a random functionality of your taste. That will certainly help you to familiarize with Phoebe plugin based architecture.
I do plan to establish a communication channel and will advise which one when set up.

[mukul-mehta]

Sure, I'll get to it. Thanks!

[yoBoyio]

@mvarlese Hello I'm interested in working on this project. Ηow can i contribute?