Technology Blogs by SAP
Learn how to extend and personalize SAP applications. Follow the SAP technology blog for insights into SAP BTP, ABAP, SAP Analytics Cloud, SAP HANA, and more.
Showing results for 
Search instead for 
Did you mean: 

SAP was invited to participate in the annual Gartner BI Bake-Off events at the Gartner Data & Analytics Summits that were set to take place last March in London and in Dallas, TX. And while the conferences didn’t physically occur, the BI Bake-Off took on a virtual life of its own in September.

We were pleased to be included in this event for BI vendors who have gained attention from the industry over the past year based on Gartner inquiry volumes.  Gartner provided the vendors with a data set from the World Health Organization (WHO) and questions to explore.

One of the beauties of working with a data set is the mysteries and stories that the data holds just waiting to be uncovered and shared.  This data set from the WHO proved to be no different.

We chose SAP Analytics Cloud to explore this data and come up with observations and recommendations. SAP Analytics Cloud is unique in that it is a true analytics product, containing business intelligence (BI), planning, and predictive capabilities—all in one integrated, seamless, end-to-end workflow.

People Friendly - Using Augmented Analytics

Augmented analytics in SAP Analytics Cloud makes data “people friendly” and enables the self-service use of analytics by anyone—regardless of their analytics skill set.

Augmented content is automatically generated as opposed to the historically time-consuming manual construction process. People focus their time utilizing machine-generated advanced insights to decide and act on the next best course of action.  This makes insight discovery faster, bringing data science to the masses, reduces human bias, and ensuring decision makers can act with confidence.

We applied SAP Analytics Cloud augmented analytics to the WHO data set. Here’s some of the interesting things that we learned.

The Big Question: What Are the Different Factors that Most Influence Life Expectancy?

Here’s a snapshot of our SAP Analytics Cloud story based on the data set provided and generated using augmented analytics.

We explored specifics for deeper information and understanding regarding life expectancy.  The chart below shows the top 15 countries and an absolute, as well as percentage variance allowing you to quickly identify those countries that have gained the most/least in life expectancy over the last 20 years. Developing countries, including several in Africa, are making significant strides.

We can see that Japan has the highest life expectancy and countries in Africa have gained far more years compared to countries in Europe and the Americas.

Looking at healthy life expectancy (HALE), we can see a similar picture. Overall, life expectancy has been getting longer, but the HALE has not gained as much as life expectancy. So, it appears that we are getting older, but we are not necessarily living a healthy life in our last years.

Uncovering Key Influencers Using Smart Discovery

Smart Discovery is a feature in SAP Analytics Cloud that uncovers new or unknown relationships between columns within the dataset to help you understand the main drivers behind the core KPIs.

A machine learning model trained on history can generate a story up to four pages long that is specific to your dataset. The model typically delivers an overview, a list of key influencers, a table of unexpected values, and an interactive what-if simulation. With Smart Discovery, you let your dataset speak for itself and you reduce the unavoidable human bias that can obscure important truths. You can investigate correlations and trade-offs among various dimensions in your data and reveal the statistics behind the drivers of your KPIs at a new level of sophistication.

The key influencer chart shows how our Smart Discovery capabilities are able to run through the WHO data set and identify the key influencing factor for life expectancy. We can see that we have two main categories of influencing factors: environmental factors, such as unsafe water and poor sanitation, and more personal influencing factors, such as diet, obesity, high blood pressure, and drug usage.

Exploring further, we can see in the popup correlated factors that are related to the one that we selected—in this example, “Unsafe water source”. So not only does Smart Discovery show us the key influencer, but it also allows us to take a deeper look at even more related influencing factors.

To better understand the details behind this trend, we took a look at further information available to us on infant mortality and under-five deaths.

We can see that infant mortality has improved greatly. Examining under-five deaths, we can see that India is the top contributor from the developing countries.

Gaining a Clearer Understanding Using Smart Insights

Next, we wanted to understand what is driving the mortality rate of young children in one of the developed countries, so we decided to look closer at the more than 50,000 deaths in the US with Smart Insights.

Smart Insights allows you to quickly develop a clear understanding of complex aspects of your business data, by letting you see more information about a particular data point in your visualization or table and explaining these findings with natural language generation in SAP Analytics Cloud.

With the click of a button, specialized algorithms run in the background to analyze all the data relevant to the information you have selected in its current context.

Using Smart Insight, we can quickly recognize some trends impacting US deaths and we can see that three out of the top 10 causes for child death are related to neonatal situations.

In real time, the system determines the top contributors and shows us that neonatal issues deserve greater analysis. Smart Insights provided us with information on the top contributors to this disturbing number. Some of which are purely physiologically related, while others involve external contributors.

Economic Influence on Life Expectancy

After we reviewed the health, personal, and environmental factors, we also needed to take a look at the financial relationship between life expectancy and the expenditure on health care.

We used a Tree Map to see the relationship in the data for financial impact on life expectancy. Smart Insight has the system telling us that the US is spending 300 percent above average on healthcare.

We can see in the Tree Map that the US is the country with the highest spending on healthcare, but we can also see that other countries—such as Japan—spend a lot less money and are even achieving a higher life expectancy.

We can also see that the spending of the US is increasing year over year, but that the life expectancy is plateauing, reflecting a “diminishing returns” situation since spending is increasing but the increase in life expectancy is very small.  Perhaps you can’t spend your way to a healthier life expectancy.

Understanding Where Life Expectancy Is Headed Using Smart Predict

In addition to the simple and immediate visual forecast, we also used Smart Predict to create a more sophisticated forecast, and to look at a scenario where we restrict the influencing factors to only the top 10.

Smart Predict is another feature in SAP Analytics Cloud that helps people of any analytics skill set answer questions that need predictions or predictive forecasts to plan for the future. Proven machine learning algorithms learn from historical data and find the best relationships or patterns of behavior to easily generate predictions for future events, values, and trends.

When we look at the top 10 key influencers for life expectancy and focus on personal lifestyle factors and discount socio-economic factors, we can see an interesting pattern. In the less developed world, life expectancy is predicted to rise; however, in the developed Western world, there is a predicted drop.  (It’s important to stress that this is based on the restricted factors, because in reality life expectancy rises everywhere).

We also did a more detailed prediction of infant mortality for each region (based on data from 1950 to 2018), and a forecast though 2032 reveals that the future looks positive here.

“What If” We Make Some Changes Using the Value Driver Tree

Here’s another way of viewing “what-if” scenarios that shows the integrated workflow and linked analysis of the analytics designer and planning capabilities, both included in SAP Analytics Cloud.

We created the value driver tree based on the key influencing factors and the data from 2016 and we organized the individual factors into four overall categories: finances, public health, personal health, environment.

On the right you see two KPI tiles, one for alcohol use and one for smoking, and a QR code to scan. In a live interactive scenario, the audience voted by QR code using their phone to see the results of a 15% reduction in smoking.

The value driver tree below reflects that change—both in the smoking tile with a reduction in the number of deaths, and the increase in overall years reflected in life expectancy at birth, as well as the delta change resulting from the input.

In this case, the Value Driver Tree analysis reflects that by reducing smoking, you can potentially add some years to your life expectancy.

Learning Today for a Better Tomorrow

The beauty of this opportunity is that it shows the versatility and value of analytics.  It reflects the opportunity to analyze, understand, interpret, and improve.

We learned that both social-economic environmental factors as well as personal lifestyle factors play a major role in impacting and extending healthy life expectancy. Governments, businesses, and individuals have an opportunity to make positive contributions here for betterment and improvement of all, both in developing and developed countries.

How much more appropriate could this learning be for all of us given the unique and challenging year that is 2020?

To try SAP Analytics Cloud for yourself, register our SAP Analytics Cloud free 90-day trial now.