Presentation given at the Canadian Institute of Actuaries Annual Meeting in June 2013. Covers the direction business intelligence is moving in for insurance.
7. Business Intelligence
Business intelligence (BI) is a set of theories, methodologies,
processes, architectures, and technologies that transform
raw data into meaningful and useful information for business purposes. BI can
handle large amounts of information to help identify and develop new
opportunities. Making use of new opportunities and implementing an
effective strategy can provide a competitive market advantage and long-term
stability.
BI technologies provide historical, current and predictive views of business
operations. Common functions of business intelligence technologies are
reporting, online analytical processing, analytics, data mining, process mining,
complex event processing, business performance management,
benchmarking, text mining, predictive analytics and prescriptive analytics.
Though the term business intelligence is sometimes a synonym for
competitive intelligence (because they both support decision making), BI uses
technologies, processes, and applications to analyze mostly internal,
structured data and business processes while competitive intelligence
gathers, analyzes and disseminates information with a topical focus on
company competitors. If understood broadly, business intelligence can
include the subset of competitive intelligence
9. Data
Warehouse
Extraction, Transformation
and Loading (ETL)
Metadata
(data about data) Online Analytical
Processing (OLAP)
Source Systems
End User
(there are other forms)
Typical BI StructureTypical BI Architecture
12. Business Intelligence
you don’t know what you don’t know
there may not be a single version of the truth
retrospective changes messy if possible at all
designed seriatim aggregations
hard to keep up to date
15. Big Data
Big data is a collection of data sets so large and complex that it becomes difficult
to process using on-hand database management tools or traditional data processing
applications. The challenges include capture, curation, storage, search, sharing, transfer,
analysis, and visualization. The trend to larger data sets is due to the additional information derivable from
analysis of a single large set of related data, as compared to separate smaller sets with the same total
amount of data, allowing correlations to be found to "spot business trends, determine quality of research,
prevent diseases, link legal citations, combat crime, and determine real-time roadway traffic conditions.
As of 2012, limits on the size of data sets that are feasible to process in a reasonable amount of time were on
the order of exabytes of data.[ Scientists regularly encounter limitations due to large data sets in many areas,
including meteorology, genomics, connectomics, complex physics simulations, and biological and
environmental research. The limitations also affect Internet search, finance and business informatics. Data
sets grow in size in part because they are increasingly being gathered by ubiquitous information-sensing
mobile devices, aerial sensory technologies (remote sensing), software logs, cameras, microphones, radio-
frequency identification readers, and wireless sensor networks. The world's technological per-capita capacity
to store information has roughly doubled every 40 months since the 1980s; as of 2012, every day
2.5 quintillion (2.5×1018) bytes of data were created. The challenge for large enterprises is determining who
should own big data initiatives that straddle the entire organization.
Big data is difficult to work with using most relational database management systems and desktop statistics
and visualization packages, requiring instead "massively parallel software running on tens, hundreds, or even
thousands of servers”. What is considered "big data" varies depending on the capabilities of the organization
managing the set, and on the capabilities of the applications that are traditionally used to process and
analyze the data set in its domain. "For some organizations, facing hundreds of gigabytes of data for the first
time may trigger a need to reconsider data management options. For others, it may take tens or hundreds of
terabytes before data size becomes a significant consideration."
19. Descriptive Title
Quantitative
Sophistication/Numeracy
Sample Roles
Data Scientist or
Quantitative Analyst
Advanced Math/Stat
Internal expert in statistical and
mathematical modelling and
development, with solid business
domain knowledge.
Business Intelligence /
Operational Analytics
Good business domain,
background in statistics
optional
Running and managing analytical
models. Application of traditional
methods such as experience studies.
Business Intelligence/ Reporting
Data and numbers oriented,
but no special advanced
statistical skills
Reporting, dashboard, OLAP and
visualization, some design, posterior
analysis of results from quantitative
methods. Spreadsheets, “business
discovery tools”
Analytic Types
Types of Analysis
Type V
20. Data Scientist Job Description
• Passion for “playing” with tons of data and supporting scientific experiments to
validate the performance of algorithms
• Advanced degree in Statistics or related area
• Experience with traditional as well as modern statistical learning techniques,
including: Support Vector Machines; Regularization Techniques; Boosting, Random
Forests, and other Ensemble Methods.
• Strong computer science skills with high-level languages, such as R, Python, Perl,
Ruby, Scala or similar scripting languages.
• Experience with Hadoop and working with multi-terabyte systems.
• Extensive hands on experience working with very large data sets, including
statistical analyses, data visualization, data mining, and data
cleansing/transformation.
• Business expertise and entrepreneurial inclination to discover novel opportunities
for applying analytical techniques to business/scientific problems across the
company.
• Good communication ability
24. Opportunities
Actuaries already have:
Most the statistical skills
Some computer skills
Business expertise
Communication skills
Significant job growth in analytics is predicted
Reputation for actuaries in analytics can be enhanced
However… we have competition
25. U.S. Department of Labor
occupations forecasted for growth in analytics
Job Titles Expected
Growth by
2018
Total #
Expected
Projected
Median
Income
Top 10%
Income
Librarians 8% 172,400 $52,530 $81,130
Accountants/Auditors 22% 1,570,000 $59,430 $102,380
Statisticians 13% 25,500 $72,610 $117,190
Ops Research Analysts 22% 76,900 $69,000 $118,130
Management Analysts 24% 925,200 $73,570 $133,850
Actuaries 21% 23,900 $84,810 >$160,780
27. “Required” Skills/Techniques
Traditional Statistical Techniques
Ordinary Least Squares
Logistic Regression
Generalized Linear Model
Time Series
Methods That Group/Organize
Trees/Clustering
Prep for Analysis
Model Validation
Conceived by Dan Bricklin, refined by Bob Frankston, developed by their company Software Arts,[1] and distributed by Personal Software in 1979 (later named VisiCorp) for the Apple II computer, it propelled the Apple from being a hobbyist's toy to a useful tool for business,[3] two years before the introduction of the IBM PC.1-2-3 was released on January 26, 1983 and immediately overtook Visicalc in sales.Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) in November 1987. Lotus was slow to bring 1-2-3 to Windows and by 1988 Excel had started to outsell 1-2-3On May 14, IBM quietly announced the end of the road for 1-2-3, along with Lotus Organizer and the Lotus SmartSuite office suite. Lotus 1-2-3's day is done.
Conceived by Dan Bricklin, refined by Bob Frankston, developed by their company Software Arts,[1] and distributed by Personal Software in 1979 (later named VisiCorp) for the Apple II computer, it propelled the Apple from being a hobbyist's toy to a useful tool for business,[3] two years before the introduction of the IBM PC.1-2-3 was released on January 26, 1983 and immediately overtook Visicalc in sales.Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) in November 1987. Lotus was slow to bring 1-2-3 to Windows and by 1988 Excel had started to outsell 1-2-3On May 14, IBM quietly announced the end of the road for 1-2-3, along with Lotus Organizer and the Lotus SmartSuite office suite. Lotus 1-2-3's day is done.
Conceived by Dan Bricklin, refined by Bob Frankston, developed by their company Software Arts,[1] and distributed by Personal Software in 1979 (later named VisiCorp) for the Apple II computer, it propelled the Apple from being a hobbyist's toy to a useful tool for business,[3] two years before the introduction of the IBM PC.1-2-3 was released on January 26, 1983 and immediately overtook Visicalc in sales.Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) in November 1987. Lotus was slow to bring 1-2-3 to Windows and by 1988 Excel had started to outsell 1-2-3On May 14, IBM quietly announced the end of the road for 1-2-3, along with Lotus Organizer and the Lotus SmartSuite office suite. Lotus 1-2-3's day is done.
Conceived by Dan Bricklin, refined by Bob Frankston, developed by their company Software Arts,[1] and distributed by Personal Software in 1979 (later named VisiCorp) for the Apple II computer, it propelled the Apple from being a hobbyist's toy to a useful tool for business,[3] two years before the introduction of the IBM PC.1-2-3 was released on January 26, 1983 and immediately overtook Visicalc in sales.Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) in November 1987. Lotus was slow to bring 1-2-3 to Windows and by 1988 Excel had started to outsell 1-2-3On May 14, IBM quietly announced the end of the road for 1-2-3, along with Lotus Organizer and the Lotus SmartSuite office suite. Lotus 1-2-3's day is done.
Conceived by Dan Bricklin, refined by Bob Frankston, developed by their company Software Arts,[1] and distributed by Personal Software in 1979 (later named VisiCorp) for the Apple II computer, it propelled the Apple from being a hobbyist's toy to a useful tool for business,[3] two years before the introduction of the IBM PC.1-2-3 was released on January 26, 1983 and immediately overtook Visicalc in sales.Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) in November 1987. Lotus was slow to bring 1-2-3 to Windows and by 1988 Excel had started to outsell 1-2-3On May 14, IBM quietly announced the end of the road for 1-2-3, along with Lotus Organizer and the Lotus SmartSuite office suite. Lotus 1-2-3's day is done.
Alternatives
This is an oxymoron, but an early driver of BI was to establish a single set of data that would enable analytics.
can’t handle stochastic modelsInflexibleand it’s still just internal data(and just reporting)
Conceived by Dan Bricklin, refined by Bob Frankston, developed by their company Software Arts,[1] and distributed by Personal Software in 1979 (later named VisiCorp) for the Apple II computer, it propelled the Apple from being a hobbyist's toy to a useful tool for business,[3] two years before the introduction of the IBM PC.1-2-3 was released on January 26, 1983 and immediately overtook Visicalc in sales.Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) in November 1987. Lotus was slow to bring 1-2-3 to Windows and by 1988 Excel had started to outsell 1-2-3On May 14, IBM quietly announced the end of the road for 1-2-3, along with Lotus Organizer and the Lotus SmartSuite office suite. Lotus 1-2-3's day is done.
Data from GartnerBig data this year will account for US$28 billion of IT spending worldwide, which will increase to US$34 billion in 2013, according to Gartner.In a report released Wednesday, the market research firm said much of 2012 expenditure will be in adapting traditional tools to address issues related to the big data phenomenon such as machine data, social data, and the large variety and velocity of data. In contrast, only US$4.3 billion will be focused on new big data functionalities.Specifically, social network and content analysis had the most impact on big data budgets, and projected to account for 45 percent of new IT spending each year, Gartner said. Application infrastructure and middleware would account for 10 percent of yearly spend.Big data is not a distinct, standalone market, said Mark Beyer, research vice president at Gartner. Rather it represents an industry-wide market force addressed in products, practices and solution delivery, he explained.In 2011, big data was the new driver in almost every category of IT spending. Through to 2018, however, big data requirements will gradually evolve from being a differentiator to "table stakes" in information management, Beyer said.Elaborating, he said by 2020 big data features and functionalities will be non-differentiating and routinely expected from traditional enterprise vendors as part of their product offerings.By the end of 2015, Gartner said it expects leading organizations to begin using their big data knowledge in "an almost embedded form in their architectures and practices". And around the start of 2018, the distinction--and advantage--new big data products had over traditional offerings that provide additional functions to handle big data, will decrease.Skills, tools and practices leading companies acquired over the years to handle big data would eventually become routine flexibility, it added.Beyer said: "Because big data's effects are pervasive, big data will evolve to become a standardized requirement in leading information architectural practices, forcing older practices and technologies into early obsolescence."In other words, big data will end up as "just data" once again by 2020, and approaches toward architecture, infrastructure, hardware, and software that do not adapt to this "new normal" will be retired, he said.
How do they move up? Type Shifting: How Analysts can slip up to higher types and what organizations need to do to facilitate itThe go-to BI analysts are ready to move up to Type IIIType III analysts can be trained to be Type IIInternal training and mentoringExternal certification programs such as one being offered by the Society of Actuaries
Term invented by YahooWho is this data scientist. There is some confusion over the term. Some define it as just skilled in statistics' and programming, others include ability to communicate with the business as a result of having domain expertise. For the most part, the “data scientist” will probably be more a collaborative group than an individual. Exception: actuaries, who have always been data scientists, but many rise to senior position in the business. Domain expertise a prerequisite for fellowship. The training of actuaries is a good model for data scientists. The value in big data is analytics. Because, as we said, The data doesn’t speak for itself. But lets take a closer look at analyticsUsed to be called them quantsFew and far between
Team Members: Lisa Tourville (Chair) Joan Barrett Guillaume Briere-Giroux Jack Bruner Kara Clark Ian Duncan Kim DwornickAlice Kroll John Lloyd David McleroyKevin Pledge Jacques RiouxChris Stehno
Significant job growth in analytics is predicted, including management and leadership roles.
Meg – why we see opportunities for actuaries…The US Department of Labor predicts healthy growth in occupations that work in analytics. In addition, McKinsey (May 2011) predicts 1.5 million managers and advanced analysts needed by 2018. A recent review of open actuarial positions with a recruiter showed 15% of all openings had a preference for job seekers with proficiencies in one or more aspects of advanced business analytics, most often predictive modeling.But a survey (Fall 2011) of 55 life insurance companies by the SOA indicated life insurance companies were not using predictive modeling techniques in any widespread way but planned to in the near future…40% considering using predictive modeling to enhance sales and marketing50% considering using for underwriting1 company currently using predictive modeling for claimsKey Points:Actuaries are included in the job titles associated with job growth in analytics. There are many job titles associated with this growth (IT jobs, too).According to this chart, actuaries tend to be more highly compensated than our fellow professionals. Our stories from actual hiring managers conflict on this. Some employers pay non actuaries at the same rate. Some employers say they are hiring economists and “ninja librarians” because they cost less.
Actuaries as leaders, not the people running the models.
Many other skills and techniques were identified as part of our study but these made the top of the list. (Others – Survival/failure Time Analysis, Factor Analysis,
Technical sites like stack overflow
R is a free software environment for statistical computing and graphics. It compiles and runs on a wide variety of UNIX platforms, Windows and MacOS.