WHY IT MATTERS: Digital Transformation

An explanation of some key mathematical concepts required to make deep learning possible.

WHY IS THIS IMPORTANT

Behind what looks like "magic" on Google search, google translate or IBM Watson lies mathematics algorithm that process large amount of data. This is essential information to understand how digital computing, big data and deep learning works.

Basically it confirms that 8hrs of sleep and weekends make you happier.

WHY THIS IS IMPORTANT?

Big data will provide us with more and more of these proofs based on facts gathered from millions of people and events.

Analysis of satellite images using Big Data tools can measure economic activity in real-time with little bias.

WHY THIS IS IMPORTANT

Companies can use this as early warning signal of growth or slowdown in the market or with competitors and take appropriate response: buying or dumping stock, ramping up or slowing down production, selecting areas to open new stores or focus marketing dollars.

Insurance company offers bluetooth toothbrushes to clients to reduce their premiums - if they brush well.

WHY THIS IS IMPORTANT

Big data and digital connected things have the potential to transform entire industries and it looks like the insurance industry is one of them. With Desjardins ajusto mobile app (http://www.desjardins.com/ca/personal/insurance/car-insurance/ajusto/) good drivers get up to 25% premium discounts, Oscar fitbit tracker users get lower premiums as well (and the startup gets a crazy 15B$ valuation http://www.businessinsider.com/oscar-raises-145-million-at-a-15-billion-valuation-2015-4), the trend appears to be clear: give us access to your data and we'll give you a better deal.

A real big data company, dataminr goes through twitter and many other sources to perform real-time information discovery.

WHY THIS IS IMPORTANT

This solution goes beyond making nice charts and delivers predictions based on a myriad of sources. It may be the tip of the iceberg to provide companies with early warming signs of impending doom or success.

Google trends now an even more useful tool for insights with real-time updates and other improvements.

In one diagram, BCG captures the impact of digital transformation: everything is being digitized and what is not already will soon be. Question is: what will your company do with all that data?

WHY THIS IS IMPORTANT

Companies that embrace change early become market leaders: there is no prize for 2nd place anymore. BCG suggest implications for executives.

Data gathered with a simple device - the withings connected scale - can provide evidence of obesity and where it is most prevalent in Europe.

No surprise in the data: people go to bed late on xmas and new year and wake up later the next morning. But the difference in each country seems to demonstrate that Americans don't party a lot and that French party more than many others. Here again, no surprise... ;-)

how grocery store food basket analysis provides new big data insights into customer segments. It leads to customer segments such as "natural sweetners".

Amazing insight on how new companies are using data to define their business model and evolve it.

Deep dive into the world of retail analytics where buying behaviour is dissected and analyzed. Provides an example where Target was able to successfully identify that a teenager was pregnant before she told anyone. Which led her parents to discover the truth when they started received coupons for baby care products in the mail. Spooky but considering what I know and that the article was written 3 years ago, this is, unfortunately, old news with old technology. Today, we are much better at it...

Must read.

How an analysis of 100K+ jawbone users tells us about the impact of weather on exercice. Also a good example of Big Data use.

Stil wondering what Big Data is all about? Well here is a good example, coming straight from Google. It shows the kind of analysis one can perform when looking at millions of google searches focussed around the house hunting theme.

It can be a good indicator of market interest - and sometimes even provide early indicators as is the case with flu trends can be predicted by Google earlier than the CDC with as much precision.

The data can also be used to analyze differences in behaviour between age groups, regions, and other criteria. This is only possible because of the large dataset as it provides sufficient data to see meaningful patterns emerge. It can even show correlations that may be meaningful in related fields, here between real estate and interior design.

Also worth reading to understand the power of BigData:

- jawbone insight into recent earthquake: How did the Napa earthquake affect sleep? http://sco.lt/5S733Z

- Bedtime Stories: Learning from Sleep Data- from #jawbone #StrataConf 2014 via @mrogati http://sco.lt/98dV8D

- google flu trends http://sco.lt/5x6g4n

This is just the beginning of insights we will get as we analyze the vast amount of data now being produced by digital devices, large or small, fixed or mobile.