The rapid growth in the amount of available data has made manual dives into large data sets a time-consuming task.

Customer relationship management (CRM) platforms, such as Salesforce, have made compiling and viewing data much easier, but spotting every trend or testing every strategy can take huge numbers of man-hours. This was the impetus for the creation of Einstein Analytics. The San Francisco-based cloud computing company has developed a new analytics program that uses artificial intelligence to quickly analyze billions of data points and combinations.

Einstein Analytics is changing how many companies approach analytics offering insight into the direction in which business intelligence is headed.

Analytics Apps

Essentially, Einstein Analytics researches the data compiled by companies, identifies trends within the data and uses this information to make recommendations for future strategy. The program is built on top of Salesforce’s CRM system.

Einstein Analytics gives users access to a portfolio of analytics apps that can be applied to data in the Salesforce Marketing, Services and Sales clouds. It is paired with Einstein Discovery, which focuses on recommendations from actionable data.

The idea is to give organizational leaders more insight into what their own data can tell them. “We have more customer data than ever before, and we need AI to turn data into something actionable for the business user,” said Arijit Sengupta, head of Einstein Discovery, in a 2017 article by Larry Dignan of ZDNet.

Ketan Karkhanis, general manager, Salesforce Analytics, provided CMSWire with an example of how Einstein Analytics might be used in a real scenario. A salesperson concerned about making quarterly numbers could use the smart cloud analytics app to dig into data on customer, competitor and sales pipeline data. Einstein Analytics could then recommend setting up a meeting with a certain customer because that person tends to close sales faster.

How It Affects Businesses

Businesses can gain an advantage by leveraging the Einstein systems in many ways, including the following:

Speed

In addition to the ability to deliver insights from data, Einstein Analytics is faster and more efficient, as Salesforce notes, than a team of data scientists. By leveraging artificial intelligence, Salesforce predicts companies will make decisions 38% faster.

Predictions

In data analytic platforms like Einstein Analytics, AI handles that task of validating every piece of data. The system also looks into customer sentiment in a variety of ways, including comments made on social media. As described in the earlier example about the salesperson, it can also be used to manage day-to-day workflow, allowing employees to optimize their time.

The Community Cloud

In the same 2017 article by ZDNet, Sengupta, head of Einstein Discovery, said, “Einstein Analytics can discover relevant content on social media, alerting companies to good content and also to influencers who write within a certain niche. This can give organizational leaders real-time feedback on certain products or services, and not just their own but those of competitors.”

Data Analysis

Einstein can crunch millions of data points on any number of “triggering events,” such as transactions that are part of daily workflow like incoming calls or even a sale. It can also analyze data from the “internet of things”, another growing area for many businesses.

Does this make data scientists obsolete? That is unlikely to happen any time soon, as business leaders tend to want decisions based on data to be made by people. However, as self-learning networks continue to expand and become more sophisticated, recommendations on business actions from AI such as Einstein Analytics could become the new normal.

A key area is readmissions, which has become a big issue for hospitals. About 20% of discharged elderly patients end up returning to the hospital in 30 days or less. Overall, about a third of all discharged patients return to the hospital in 90 days.

Not only does readmissions impact the health of patients, but it also raises costs for the hospital, patient, insurance companies and taxpayers who are funding care through government programs.

However, predictive analytics is offering medical operations a way to lower readmission rates and ensure their patients are receiving the highest quality of care.

Facing Fines for Readmissions

Part of the catalyst for addressing the issue is that hospitals now face penalties from Medicare for readmitting certain patients in less than 30 days. These cases include certain types of lung ailments as well as heart attacks, heart failure, pneumonia and hip and knee surgery.

The fines can reach as high as 3% of total Medicare payments. According to the U.S. Centers for Medicare and Medicaid Services, 2,655 out of a possible 3,454 hospitals were penalized in 2015. About 15% of those received 1% penalties or more, while 1% of hospitals were penalized 3%.

To lower penalties, medical facilities are taking steps to ensure patients receive proper care after discharge. For example, some issue medications to patients that they could not typically afford, or work to have medical personnel outside the hospital monitor a patient’s progress.

Predictive analytics is also playing an increasingly important role. By taking a closer look at patient data, hospitals can better identify those with a high risk of readmission. Factors include:

• Number of hospital visits in the previous six months
• Length of the original stay
• Acuity of the medical issue at admission
• Socioeconomic factors
• Past drug use

Leveraging Data Through Risk Score Reporting

Using patient data to identify those with a higher risk of readmission is known as risk score reporting, and gives hospital personnel insight into which patients need immediate attention. In this case, predictive analytics is helping hospitals simplify their workflows and allow staff to focus on patients that are truly high risk. Some of the questions asked during the creation of the risk report include:

• Does the patient have any problems with medications or are they on any high-risk medications?
• Does the patient have any psychological problems (e.g. depression, anxiety or substance abuse)?
• Does the patient have a principal diagnosis such as cancer, diabetic complications or COPD?
• Does the patient have any physical limitations?
• Is the patient able to prove they fully understand their care plan?
• Does the patient have a reliable caregiver to assist with care after they are discharged?
• Has the patient had any unplanned hospitalizations within the last six months?

Effective risk score reporting is dependent on timeliness and relevance. It’s only useful if hospital staff act quickly on data that best reflects the patient’s current situation.

To help with this, some hospitals have set up transition teams that work with patients leaving the hospital to ensure they receive proper care.

Case Study – Allina Health

Allina Health operates a not-for-profit healthcare system through Minnesota and Wisconsin that includes 13 hospitals, 90 clinics and 16 pharmacies.

Company executives adopted predictive analytics to combat high readmission rates. They decided to focus on the first seven days after hospital discharge after reviewing statistics that showed about 40% of readmissions happened in this timeframe.

One issue they discovered was that medical staff in different locations or clinical specialties used different criteria for determining if a patient was a high risk for readmission. To fix this problem, they designed a more standardized approach and set up a regular meeting between the high-risk patient, the hospital transition team, the family and the in-home health providers to determine what type of care the patient needed after discharge.

They also used predictive analytics to incorporate information from electronic health records to determine a patient’s probability of readmission, with the highest score having a 20% or greater chance of being readmitted within a month.

The result was a 10.3% reduction in patient readmissions, including a 21% reduction among moderate to high-risk patients. They also saved an estimated $3.7 million in costs due to fewer readmissions.

By taking time to educate staff, patients and all other stakeholders on how the process works, Allina providers were able to set goals for cutting down costs and reducing readmissions while providing better care for patients.

More than two dozen states offer funding based on how many students an institution graduates, not how many it enrolls. The cost of recruiting and educating students also is continuing to rise, making student retention even more crucial to the bottom line. As costs increase, colleges hope to overcome this challenge by balancing the resources they use to recruit students with revenue generated when those students are retained.

To help achieve this, institutions are using predictive analytics to analyze demographic and performance data to predict whether a student will enroll at an institution, stay on track in their courses or require support so they don’t fall behind.

Using predictive analytics in this way makes it possible for institutions to meet their annual enrollment and revenue goals with more targeted recruiting and strategic use of financial aid. Additionally, predictive analytics allows colleges to tailor their advising services and personalize learning to help improve student outcomes.

Predictive Analytics in Higher Ed

Institutions in higher education are using predictive analytics as a way to respond to the many business and operational changes happening in the education industry. Below are the three main reasons colleges are using this tool:

Targeted Student Advising

Few colleges have an adequate number of advisors on staff, and as a result, students often cannot receive the individualized attention they need. A survey conducted by the National Academic Advising Association (NACADA) found that the national average caseload of advisees per full-time professional academic advisor was 296-to-1. The ratio jumped to 441-to-1 at community colleges.

However, systems based on predictive analytics like early-alert and program recommender can help identify students who are in need of support and allow staff and faculty to assist.

Adaptive Learning

Colleges are also using predictive analytics to develop adaptive learning courseware, which is designed to modify a student’s learning route based on their interactions with the technology. Using predictive analytics in adaptive learning platforms can help instructors pinpoint students’ learning gaps and then customize the academic experience so it better aligns with how students learn.

This tool helps students accelerate their learning by allowing them to quickly go through content they already know, while providing additional support in areas where they struggle.

Manage Enrollment

Colleges are also using predictive analytics to better inform enrollment management plans. This information helps schools forecast the size of incoming and returning classes. It is used to help the school narrow the focus of their recruitment and marketing efforts so they are only targeting students who are most likely to apply, enroll and succeed. Predictive analytics helps colleges anticipate the financial need of incoming and returning classes to determine whether a student will accept the financial aid award offered to them.

Approaching Predictive Analytics Ethically

Analyzing personal student data using predictive analytics requires careful attention to ethics and privacy.

According to The Atlantic, the best predictive models avoid making recommendations based solely on a student’s financial or cultural background. Because structural inequality makes up so much of the world, students with low-income backgrounds, first-generation students, and students of color tend to graduate with college degrees at much lower rates when compared to affluent white students.

Therefore, when institutions use predictive analytics to look at race, ethnicity, age, gender, or socioeconomic status to determine which students to target for enrollment or intervention, they can intentionally or unintentionally reinforce that sense of inequality.

For colleges and universities that are just learning how to use analytics to make decisions, remaining ethical in their use of student data can be a struggle, which is why a high degree of training and security is necessary.

Predictive Analytics Case Study

Georgia State University is known as a leader in leveraging data to provide individualized attention to students who need it most. Higher Education Marketing reports that in the last decade, Georgia State has tracked more than 140,000 student records and 2.5 million grades to identify 800 different factors that put students at risk of dropping out. Some of these risk factors include enrolling in the wrong course for their major or low grades in an introductory course needed for a particular major.

When any one of these academic mistakes occur, an alert is triggered in their early-warning system. A one-on-one student intervention is initiated within 48 hours of the alert. More than 51,000 interventions were conducted in 2016. Georgia State also added dozens of academic advisors, centralized operations and information sharing, and expanded on current resources such as peer tutoring. As a result of these changes, graduation rates have increased over the last decade by 22%, and students are completing their degrees half of a semester sooner on average.

The school has seen the most improvement with at-risk minority, first-generation and non-traditional students, who were previously falling through the cracks. The insights gained from their predictive models enabled them to anticipate students at financial risk and student demand for specific courses, which helped make scheduling processes more efficient.

Workers in the manufacturing industry are already familiar with losing their jobs to machines, but now, those working in creative fields such as marketing, writing, music and art are beginning to worry as well. In the near future, scientists may be able to build a machine that can write a book, compose a song or create a painting, leaving many who work in creative fields wondering what lies ahead.

That’s a reasonable fear for a society that has been exposed to movies like “2001: A Space Odyssey,” “The Terminator” series and, more recently, “Ex Machina.” In every one of these movies, the machines equipped with artificial intelligence (AI) eventually begin to think like humans and in the end outsmart their creators.

However, it is important to keep in mind that those plot lines are rooted in the realm of science fiction. When it comes to concerns regarding artificial intelligence and job security, it is important to know exactly what AI is, what it can do and where scientists are with its development.

Understanding AI

AI is a part of computer science. The goal for those in the field is to develop machines that can do tasks normally done by people – specifically, tasks that require intelligence.

Drilling down a bit deeper, scientists break AI into three different categories: weak AI, strong AI and the middle ground between the two. Strong AI refers to machines that can think like humans do and actually simulate human reasoning. The ability to better understand how people think would help benefit the future of AI research. Naturally, the complexities of human thought are difficult to replicate and as of yet, no machine has been built to consistently think like humans.

Weak AI seeks to mimic human behavior. For instance, IBM’s Deep Blue is a system with an expertise in the game of chess, although it does not play in the same way that humans do. These kind of machines cannot teach us anything about the way humans think, unlike strong AI systems.

The third category lies in between strong and weak AI. These systems, such as IBM’s Watson, are inspired by human reasoning, but do not simulate the actual way people think.

Plans for AI Seek to Benefit Creatives, Not Replace Them

A 2016 report published by World Economic Forum (WEF) estimates that up to 5.1 million jobs could be lost over the next five years in the world’s top 15 economies due to disruptive labor market changes such as AI. However, according to AI Business, it is more likely that people’s jobs will simply evolve so that they may work alongside AI. People will still have their jobs, but their duties and responsibilities will likely be different.

Professionals working in creative industries have a lot to gain from this technology. AI has the potential to become the ultimate creative tool, helping to build richer experiences for consumers with long-lasting value.

Design

Agencies often spend a lot of time on simple and repetitive tasks. It’s typically not the strategic thinking behind website or campaign designs that takes up valuable time, but rather the actual process of executing these ideas. The creative thinking behind these designs is purely human, but once an idea is developed, a designer may spend weeks in front of a computer bringing it to life.

This is one area where AI could prove useful. Because design is a series of rules that can be learned, in theory, AI could be taught to perform these tasks, generating websites in minutes rather than weeks or months. With AI handling the task of carrying out human plans, more time could become available for people to focus on the creative thinking behind their work.

Ideas

According to fastcocreate.com, a branch of Fast Company, there are computers currently writing book manuscripts and composing symphonies, offering proof that AI can come up with ideas on its own. The problem is, those ideas lack context and feelings. Computers don’t understand emotion or cultural relevance and they don’t get excited, angry, jealous, depressed or elated – emotions humans experience almost every day.

They also don’t invent. If you teach a computer how to write a novel, it won’t go on to produce something experimental. It will just keep writing variations of the same book over and over again.

However, that doesn’t mean AI doesn’t have anything to offer during the creative process. By performing data research, AI can help humans better understand cultural trends and interests. Yet, it’s still up to humans to translate this information creatively.

Music

In the summer of 2016, Google introduced a new group of employees dedicated to making AI more creative at a music festival known as Moogfest. The overall goal of the Google group is to see if AI can be trained to create its own art, music and video and then have it affect a group of listeners emotionally. The group’s first project is titled Magenta and it will allow researchers to import music data so the AI can be trained on musical knowledge.

The ability to think creatively has long been understood as a skill that only humans possess, so getting AI to think creatively would be one step closer to creating machines that could think on their own. While that probably won’t happen for some time, AI is still proving useful for those working in the creative sphere.

For example, when composer David Cope was hit with writer’s block, he was able to create a program that could be given various compositions and then create one in a similar style. The results helped Cope write new works. He now offers compositions by the program, called “Experiments in Musical Intelligence.” The main goal was not to create computer-generated music, but to use AI to spark new ideas.

The Future of AI in the Workforce

Ready or not, AI and robots are set to become a big part of the workforce in the very near future. According to AI Business, experts are predicting there will be limitless interaction between humans and machines, which will forever change the way humans live, work and relate to one another.

Machines are already handling much of the data collection once done by humans. Big, technology-driven companies continue to invest in AI that can crunch large amounts of data and make recommendations to consumers.

Netflix is a prime example. Their research, using data from millions of consumers, drives the company’s decisions on what shows to produce and even which cast members to hire.

Salesforce’s Einstein, one of the newest AI engines on the market, is equipped to route jobs to the right members of the sales team based on who is the best match for the customer’s needs. These two scenarios are examples of how AI enables businesses to be more efficient, accurate and effective by allowing for better use of resources, which reduces waste and improves services.

So set aside those movie-generated fears. In real life, AI is here to help.

The rise of data analytics has led to increased media attention toward event forecasting websites such as FiveThirtyEight and PredictWise, which use advanced statistical models and opinion poll analysis to predict political elections, economic trends and sporting events.

Recently, these statistics-heavy platforms were used to help predict the outcomes of national events such as Brexit and the 2016 presidential election.

Many opinion polls and data analysis surveys indicated voters would choose to stay in the European Union (EU). Likewise, polls in the U.S. predicted Hillary Clinton would defeat Donald Trump last November.

In the aftermath of those nationwide events, the reliability of big data was questioned because statistical models predicted the wrong outcomes. Or did they?

The Misinterpretation of Big Data

Collecting large sets of data has become a mainstay for many industries, but analyzing and interpreting it has been a challenge for many organizations.

In a 2015 survey conducted by PricewaterhouseCoopers (PwC), 1,800 senior executives from mid- to large-sized companies were surveyed, and results revealed that only a small percentage of companies reported effective data management practices. Of the executives surveyed, 43% said they had received few tangible benefits from big data, while 23% reported zero benefits from data analytics.

The fault, in many cases, is not with the data itself, but with the way it’s gathered, examined and translated. In terms of the predictions regarding Brexit and the 2016 presidential election, there was nothing fundamentally wrong with the data. The problem was the people in charge of collecting, analyzing and interpreting it. 

Data sets and data-driven forecasting models can often reflect their own creator’s biases. This, and the subjective interpretation of data, are examples of how data can be misunderstood.  

According to The New York Times, data experts say the danger with data analysis is trusting it without grasping its limitations and the potentially flawed assumptions of the people behind the creation of predictive models.

Big Data Provides Probabilities, Not Answers

Predictions are merely a probability something might happen, and no predictions are guarantees. The Brexit vote and the 2016 presidential election are examples of the potential shortcomings of polling, analysis and interpretation, both in how the numbers were presented and perceived by the public.

Many of the opinion polls and predictions leading up to the European Union referendum were based on telephone surveys, which indicated a majority of voters wanted to remain in the EU. The same type of opinion polls projected Trump losing the presidential election in almost every forecast model. When looking at these two cases, the collection and interpretation of data may have been flawed. Opinion polls are typically done in person and in some cases, people may not express their true beliefs and risk starting an argument or offending someone. 

In both cases, online polls provided a better indication of the eventual outcome, despite their unscientific nature. Predicting political, economic or sports outcomes is challenging to do weeks or even months ahead of time. The smallest variables can cause changes that can affect the outcome of the event.

Forecasting Big Data

Collecting data is one thing. Interpreting it and understanding the potential risks of making flawed assumptions is another.

Data analysts create methods to collect data and analyze small but important patterns that emerge from large data sets. Data analysts, also known as data scientists, can help business leaders take a more practical approach to data. This starts with knowing exactly what information is needed, rather than collecting large quantities of statistics that serve no purpose.

Bridging the gap between data scientists and those reporting the data is the logical next step to help improve event forecasting. Data scientists help provide a sound strategy for interpreting data and using it in the business. According to the PwC survey, of the 66% of respondents who reported little to no benefits of big data, a majority of them did not have a data analyst on staff. Others reported having data analysts but not utilizing them in the right context.

It’s important to understand what types of data help in making the right business decisions. Looking back at the data gathered during Brexit and the 2016 presidential election, some of that data proved to be more accurate because it came from a more reliable source.

Despite big data’s perceived failure, its usefulness in business is growing. However, employing professionals who understand how to collect the right kinds of data and interpret it correctly will be key to produce more reliable forecasting.

This advanced surveillance is driven by business intelligence: using data and patterns to make insightful and meaningful decisions. The NSA gathers multitudes of data about the American population, and its goal is to use those findings to keep the country safe from internal and external threats.

Some Americans, however, aren’t totally in support of advanced surveillance, according to research by the Pew Research Center. There’s no consensus among the public, and a measurable sector remains concerned about the lengths that the government may go, and the rights that may be violated to keep the country safe from harm.

Business Intelligence in Government Surveillance

Humans generate a lot of data. A recent study by IBM showed that we create more than 2.5 quintillion bytes of data every day. That’s more than two billion gigabytes over a 24-hour period, or the equivalent of downloading every single one of the approximately 35 million songs on iTunes 20 times.

That data is made up of photographs, emails, purchase history, text messages and more. By using business intelligence, the NSA can search for and find patterns in previously tagged data, which allows the agency to make intelligent predictions about peoples’ intentions.

Data Tagging

To identify connections or patterns through the process of data mining, metadata tags must be present. Metadata is data about data, and the label given to a specific piece of data is called a tag. When data mining, tagging data should be the first step. Without tagged data, analysts are unable to classify and organize the information so it can be processed and searched.

Tagging data also allows analysts to evaluate the information without examining the contents. This is an important point regarding the legality of data mining by the NSA, since a warrant is required to investigate the communications of U.S. citizens and lawful permanent resident aliens. However, metadata on a tag is not protected as such so analysts are able to use it to identify suspicious behavior.

Identifying Patterns

According to International Data Corporation (IDC), a data analysis firm, roughly 3% of the information that exists in the digital world is given a tag when it is created. This makes it necessary for the NSA to have a software program that puts billions of metadata markers on the information it collects. These tags serve as the foundation for any system that makes links among different kinds of data, such as video, phone records and documents.

For example, this data mining process would call attention to an individual who searches for bomb-building guides online, subscribes to terrorist propaganda and invests in explosives. That person would then be labeled as a potential risk.

But the examples aren’t always so straightforward. Even with business intelligence helping to piece together patterns and tendencies that would’ve gone previously unnoticed, the process isn’t as simple as it sounds.

Exposed Government Surveillance Programs

Two of the government’s major surveillance programs have already been exposed by former NSA employee Edward Snowden. It all began when Snowden released information about how the U.S. government was collecting phone records of every Verizon Wireless customer, which included millions of Americans. Shortly after that, two other government surveillance programs known as SKYNET and PRISM were uncovered.

• SKYNET – The NSA-created program called SKYNET monitored the location and communication patterns of people of interest by collecting mobile network metadata and bulk call records, according to an article from The Intercept. The government deployed SKYNET on a journalist named Ahmad Muaffaq Zaidan, whom they (incorrectly) suspected was a courier, and through his mobile phone usage, the program labeled him a terrorist. However, this was done without any other significant evidence. Zaidan denied the accusations and strongly criticized the invasive surveillance methods used to label him as a terrorist.

• PRISM – Established in 2007, the PRISM program was a cache of file transfers, emails, videos and other data collected from internet companies across the United States. U.S. officials claimed the program helped them catch Khalid Ouazzani, a naturalized U.S. citizen who the FBI suspected was an extremist and was planning to blow up the New York Stock Exchange. This example may be evidence that PRISM has foiled at least one terrorist plot, but at what cost?

The Cons of Mass Surveillance

The cost of American privacy is the most obvious tradeoff of surveillance programs fueled by business intelligence, but it doesn’t end there.

The NSA is, very likely, spending billions per year on monitoring the public. Economically speaking, the government’s surveillance efforts are sabotaging consumer relationships with popular brands like Verizon and Yahoo.

Furthermore, the surveillance supports the decay of internet-based securities. The NSA is collaborating with American manufacturers to design their products with features and mechanisms that make them easier to manipulate and employ as spying tools.

Business intelligence, combined with these fervent observational tactics, help keep Americans safe from threats. But, in turn, some may contend that they rob citizens of their trust and businesses of their credibility.

Is it worth it?

The country has yet to decide.

According to the Statistical Analysis System Institute (SAS), predictive analytics uses big data, statistical algorithms and machine learning techniques to predict the probability of future outcomes and trends based on historical data.

Reports have shown that businesses are spending an estimated $36 billion on the storage and infrastructure of data. Big data investments are projected to double by the end of the decade, eventually accounting for more than $72 billion, according to a 2016 report by SNS Telecom. As a result, many companies have hired data analysts and data scientists to help collect the vast amount of structured and unstructured data, compile it and then provide analysis based on that data.

Why is Predictive Analytics Important?

Traditionally, big data is compiled to understand customer habits or identify business trends. Predictive analytics, on the other hand, gives a probability for how a particular customer will behave in a future situation and how they might react to the different interactions between them and the business. Predictive analytics can help businesses discover patterns in data that can help expose problems and identify opportunities for growth.

According to the SAS Institute, some of the most common uses of predictive analytics in business include:

• Detecting fraud
• Optimizing marketing campaigns
• Improving operations
• Reducing risk

How Does Predictive Analytics Work?

As big data continues to grow in demand, it is important for professionals, especially those without an understanding of data science or business analysis, to learn the basics of predictive analytics technology and how it works.

According to the Harvard Business Review, successful predictive analytics strategies need three things.

1. Data – The most common barrier faced by organizations trying to implement predictive analytics is a lack of reliable data.
2. Statistics – Regression analysis, which estimates relationships among different variables, is the primary tool used by organizations for predictive analytics.
3. Assumptions – Every predictive model has an assumption behind it, and it is important to know what that assumption is and monitor whether it is still true. The general assumption in predictive analytics is that the future will continue to mimic the past.

Businesses that are able to gather enough relevant data, develop the right type of statistical model and monitor their assumptions carefully will typically produce more accurate predictions of the future.

What Industries Use Predictive Analytics?

Implementing predictive analytics can help companies increase revenue, improve business processes, reduce risk and provide a forecast of future behaviors and trends in the industry. Here are a few examples of how predictive analytics principles can be applied to an industry:

Retail – Retail companies can use predictive analytics to drive predictive search to their websites and offer recommendations to their customers, according to an article by Business2Community. A key benefit of predictive analytics for retail companies is the real-time processing of past data, which makes it possible to offer customers content based on their browsing history. According to SAS, Staples, the office supply retailer, achieved a 137% ROI by using predictive analytics to better understand and serve their customers.

Healthcare – Due to the large amount of medical data and electronic health records, predictive analytics in healthcare typically involves a much larger amount of metrics and data points. However, with more data comes more opportunities to learn from that data. Healthcare professionals can use predictive analytics to analyze patient data, which can help doctors forecast the possibility of illness and help with early diagnosis. Predictive measures like this can reduce hospital re-admissions, lower healthcare costs, identify high-risk patients, reduce hospital wait times and promote healthy initiatives.

Banking and Finance – When it comes to the financial industry, there are huge amounts of data and money being put at risk on a daily basis. However, this industry is no stranger to predictive analytics. Financial institutions often use predictive models to gauge a candidate’s credit card spending, optimize risk management, detect and reduce fraud, increase customer retention and maximize selling opportunities. Overall, predictive analytics helps the banking and financial industry make customer-focused decisions, forecast the likelihood of fraudulent activity and understand and rectify customer satisfaction and complaint trends.

We’re not talking about the NFL. We’re talking about the other version of big-league football, the kind folks play on their phones and laptops using spreadsheets and data sets.

Around the world, multitudes of people are engaging in fantasy football, an analytical and numbers-driven phenomenon that allows fans to assume the roles of a team’s coach and executive board. Tasks like player evaluation, salary management, teambuilding and even game day strategy are handled by a single fan who acts as owner, architect and tactician over his or her unique team.

Finding Your Fantasy Football Edge

Like in the NFL, stakes are high in fantasy football. Many players can spend months studying websites and sports news, planning for the draft, analyzing new coaching schemes or roster moves, all to find a winning edge when the season kicks off.

That’s called using big data.

Big data consists of large and complex data sets that are used to draw useful insights about something complicated. It’s been used for things like customer buying trends, internal processes and business operations.

Football has as many data subsets as any business function. Finding your edge with big data is about targeting the best subsets and using them to make informed decisions.

Here are a few surprising insights gathered from fantasy football’s big data pool.

• Don’t worry about yards or touchdowns when you’re drafting players – This might seem like a total 180-degree violation of fantasy football bylaws, but there’s some brilliant truth to this. Yards and touchdowns are among the most difficult stats to predict, which means they’re among the most difficult stats for players to consistently replicate. Instead of studying a running back’s previous yardage and touchdown totals, focus instead on how many opportunities he’ll get to run the ball, and in what situations those opportunities will arise. Ditto receivers – instead of looking at previous catches and scores, try and discern how often a quarterback will throw in that receiver’s direction during the coming season. If you figure that out, all the other metrics will take care of themselves.
• Ignore a player’s average yards per play – It can be enticing to spend a late round draft pick on a running back who averages 5.4 yards per carry, because a player’s average yards are based on things like long runs or big losses, and there’s no consistent way to predict either of those outcomes. A runner might average 8.5 yards per carry in game one, and then 1.5 yards per carry in game two, giving him a healthy 5.8 rushing average that doesn’t tell the whole story.
• Your goal is to score points, so focus on points – The football research website Advanced Football Analytics has a metric called Expected Points Added Per Play. It’s a complicated mishmash of several situational circumstances and historical trends, but it equates to the number of points a player will accumulate on any given play. For example, say a starting running back gets hurt in the NFL. You might find his replacement in your fantasy league’s free agency pool, but you notice the replacement is only averaging 3.1 yards per carry and hasn’t scored a touchdown since 2014. But, the Expected Points Added Per Play metric tells you that this replacement’s statistical output was a product of his situation (being used only to block on passing plays, for example). Now that he’s in the starting lineup, his role (and fantasy football value) is expected to change in a really positive way.

Big data is all about insight. If you know where to look, a lot of your draft and game day decisions can become a matter of math instead of instinct. So bring a different cheat sheet to your fantasy football draft this season – and win with big data.

Here we will discuss how that data may be used.

Read below to learn about eight interesting ways that big data is currently being leveraged in the business world.

1. Getting an Aerial View of Your Values

Publicly available satellite data is something that may not appear to deliver much information other than the layout of cities or terrain features.

However, some analysts argue that much can be learned about people and organizations based on what is visible from the sky.

For example, a company called HaystaqDNA has developed algorithms that can detect whether residents have solar panels installed on their roofs.

This allows the company to have a sense of which residents may be more environmentally conscious and have enough disposable income to invest in new technology.

>> Learn Information Systems Security with Villanova University

3. Predicting Employee Success

HR departments are typically interested in creating detailed profiles of their employees and trying to quantify workplace performance.

Yet, some companies specialize in deeper analysis.

For instance, big data company Evolv suggests that even the web browser a person chooses to use can say something about future performance.

Data suggests that users who use alternative browsers like Google Chrome or Mozilla Firefox often show higher job performance than people who use the default browsers packaged with their systems.

5. Discovering Lifestyle Changes

Purchase data can do more than allow companies to focus on certain advertisements.

In one case, Target was able to use big data to discover that one of its customers was pregnant, and later sent her coupons for baby products.

Statistician Andrew Pole developed a formula comprised of 25 products that determined the likelihood of pregnancy when the products were purchased together.

Individuals who unknowingly participated in the experiment did not know they were pregnant at the time of the experiment, but later confirmed Pole’s system was correct.

7. Crawling Email for Interests

Many companies often have an active interest in the content of a user’s emails.

Suppose you send an email to a friend about a pet using a popular service like Google’s Gmail.

Afterwards, when you see advertisements about animal products, both within Gmail and throughout other areas of Google, this is often a result of big data analysis.

Google Ads uses algorithms to scan emails and deliver targeted advertisements depending on common user topics.

Big Data Legality

The insights provided by big data can be extensive, with companies continuing to innovate and develop new ways of understanding their customers based on the increasing amount of data available to them.

However, some privacy advocates may question the legitimacy of using big data. The final article in this series explains how companies can legally collect and use your data.