# ðŸ’¦ Towels have a dry sense of humor

Late to the Party ðŸŽ‰ is about insights into real-world AI without the hype.

Hello internet,

Can you believe that March is over already? What a crazy Q1 already with chatGPT and the whole GraphCast craziness. Sure keeps life interesting. Letâ€™s look at some fun things I found this week!

## The Latest Fashion

- Enrich your NLP datasets with extras like link extraction or sentence complexity
- This library helps you refurbish your Python code to best practices with explanations
- You can make 3blue1brown-style videos for machine learning now!

*Got this from a friend? Subscribe here!*

## My Current Obsession

I worked on PythoDeadlin.es and overhauled the codebase. Now every conference has its own proper page, and I even added maps to each page and a big one here: pythondeadlin.es/map. Honestly, really proud of this one. Let me know if you have more features you think might be useful!

A few new people checked out the Latent Space, my free inclusive ML community. Considering how Twitter is going currently, I feel reinforced by my decision to slowly get this started!

## Thing I Like

I upgraded my vacuum robot to a Roborock S7 MaxV. Super fancy. Super fun. Unfortunately, I am now spending extra time just watching it scrub my bathroom tiles because itâ€™s so satisfying.

## Hot off the Press

I created more 1-minute videos about data science and machine learning concepts!

- How AI is used every day
- Classification vs Regression in 56 seconds
- The best way to do machine learning on Excel sheets
- Get your expert knowledge into ML models
- Start doing data augmentation

Itâ€™s been really fun. What do you think? Should I do more of these?

### In Case You Missed It

Do not include these data science projects in your CV

## Machine Learning Insights

Last week I asked, ** What is correlation?**, and hereâ€™s the gist of it:

Correlation is used to quantify how much two variables are linearly related.

Correlation is measured on a scale from -1 to 1, where -1 represents a perfect negative correlation (i.e., when one variable increases, the other decreases) and 1 represents a perfect positive correlation (i.e., when one variable increases, the other also increases). A correlation value of 0 indicates no relationship between the two variables.

For example, let’s consider meteorological data on temperature and humidity. We might want to investigate whether there is a correlation between these two variables. A positive correlation between temperature and humidity would mean that as temperature increases, so does the humidity. A negative correlation, on the other hand, would indicate that as temperature increases, humidity decreases.

We can use a correlation coefficient, such as the standard Pearson’s correlation coefficient, to measure the correlation between temperature and humidity. This coefficient measures specifically the linear relationship between two variables. Suppose the correlation coefficient is close to 1 or -1. In that case, it indicates a strong linear relationship, whereas a 0 suggests they are not linearly related. This does not mean there isn’t a non-linear relation, which is vital to remember when analysing data.

## Data Stories

Have you ever tried visualising text?

Well, there are more ways than doing a word cloud. Apparently. This is a nice collection of text visualization methods, from graph-based to documents to anything else you can come up with!

There are even ways to filter it!

When you have a presentation about NLP, it can get incredibly boring without proper imagery. So definitely worth checking this one out:

Source: Textvis Browser

## Question of the Week

**What is the relation between standard deviation and variance?**

*Post them on Mastodon and Tag me. I’d love to see what you come up with. Then I can include them in the next issue!*

## Tidbits from the Web

- What is really happening in the Bermuda Triangle?
- Falsehoods programmers believe about time
- Turns out you donâ€™t need a computer to make techno

Jesper Dramsch is the creator of PythonDeadlin.es, ML.recipes, data-science-gui.de and the Latent Space Community.

I laid out my ethics including my stance on sponsorships, in case you're interested!