“We think we want information when we really want knowledge.” 

by Nate Silver 

Hi there!

It’s Kan from Exploratory.

Yes, it’s been a long time for most of you!

Sorry, I haven’t been able to keep up with this weekly update email!

Anyway, here it is. Hoping to get back on track.

I have one news for you if you’re an Exploratory user. Please read ‘What We Are Working’ section below.

Now, here’s this week’s update!

What We Are Reading

3 ways to make better decisions using the power of noticing - Link

I didn’t know about this book called “The Power of Noticing” until recently. But after reading this introductory essay about the book, now I have to read it.

One of the points in the essay is about noticing something that is not there - “A dog that didn’t bark”. This is super critical for any data analysis works especially when it is for decision making.

P values are just the tip of the iceberg - Link

Jeffrey Leek and Roger Peng, associate professors at Johns Hopkins, argue many steps of the data analysis that are equally or even more important than P values but are not discussed as much as it is done for P values.

Leaders need to be data informed rather than being driven by data. Understanding these underlying processes of data analysis is critical as more people will make decisions based on data.

How a Feel-Good AI Story Went Wrong in Flint - Link

There is a saying of “You get what you deserve.”

Even when companies like Dell used the same creative agency that Steve Jobs’ Apple was using at the time, the ads produced for Dell was not great.

AI can be similar in that if you don’t know how AI works you won’t be able to have a great AI system that would benefit you.

And that’s basically what happens in Flint, Michigan. AI system was built to detect contaminated water pipes that could kill people. But the mayor, contractors, and other people involved in the project killed it and ended up spending all the money for some other ways without making much progress.

I’d imagine there will be tons of these types of AI systems that are currently under development but will be forgotten because we won’t have enough people knowing what to do with them.

Artificial Intelligence Hits the Barrier of Meaning - Link

“While some people are worried about “superintelligent” A.I., the most dangerous aspect of A.I. systems is that we will trust them too much and give them too much autonomy while not being fully aware of their limitations.”

Prisons across the U.S are quietly building databases of incarcerated people’s voice prints - Link

These are the kinds of things that are happening everywhere in the world while we are focusing too much on sexy subjects like Facebook and Google. And very concerning.

What We Are Writing

We have written the following Geocoding related posts recently!

  • Geocoding US Address Data with zipcode Package & Visualize it - Link
  • Reverse Geocoding Part 1 — Using Boundary Data with GeoJSON - Link
  • Reverse Geocoding Part 2 — Using Google Maps APIs - Link

What We Are Working On

Exploratory v5.1

We have decided to delay the release of v5.1 due to some heavy lifting works that were needed to make both Exploratory Desktop and Server more robust for the parameterized reporting feature.

The good news is that we could push a few important new features into the release while we were working on the parameter feature.

The new release target date for v5.1 will be towards end of this month. If you can’t wait please contact me at (kan@exploratory.io) to get a stable version of Exploratory v5.1 Beta.

Anyway, here are a few new chart types we have added to v5.1.

Violin Plot

Violin plot is basically what you get when combining Density plot and Boxplot. It helps you understand how the data is distributed more intuitive way and compare the distribution among categories.

You will have an option of showing the boxplot and the dots (original data points) along with the violin plot.

Scatter Matrix

Finally, Scatter Matrix is in the house!

This chart will be introduced under Analytics view’s Correlation in v5.1. initially. It will make it more intuitive to understand the correlation between any given pair of numeric variables.

Heatmap - Repeat By Support

We have added ‘Repeat by’ support for Heatmap chart! You will be able to see multiple heatmap charts for multiple categories.

That’s it for this week.

Have a wonderful week!

Kan CEO/Exploratory

Subscribe Weekly Updates!

This is a weekly email update of what I have seen in Data Science / AI and thought were interesting, plus what Team Exploratory is working on.