My top 3 Data Science Hidden Gems

Disclaimer: This article is not a hidden ad for the tools, they're mainly open source

>5 min read - Here are some very handy and useful tools I used in the past but surprisingly also found are not very well known across the community, let me try to fix it with this quick introductions:

1- Facets: Explore interactively your data with Facets

Designed to be a data exploration and preparation helping tool, I also like to use it to present data:

Here is a presentation video

Here is the Reference website where you could find the "Try it out with your own data" button that allows you to upload a CSV file and explore it with Facets

2 - Tensorboard: Audit your machine learning models with Tensorboard

Take a look at your machine learning model training performances and results on a human readable way

In order to access the tensorboard, in your TF code save the training logs in a folder then use the tensorboard command to get a very delightful visualisation on what happened during the model training

Here is a Youtube presentation:

Check out the Reference website

3- Use a GPU with Collaboratory

A free Jupyter notebook that comes with the most common ML libraries pre installed is available on the web. This tool is named Colaboratory or Colab. It even let you use GPU for your model trainings. I also use Colab on my day to day job connecting directly to Google Cloud Big Query to showcase some ML models prototypes.

Youtube presentation:

You could use it directly here: Reference website

I really encourage you to try out those tools, let me know in comment if you also have other Data Science & ML Hidden Gems in mind.

Bonus 4 - Humour and ML instagram account

After or in between hard work sessions you may found funny the memes posted on this instagram account : @neuralnetmemes (i'm not the owner nor responsible of whats posted)

Have a nice day :)