O slideshow foi denunciado.
Seu SlideShare está sendo baixado. ×

How relevant is Predictive Analytics relevant today?

Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
Anúncio
How relevant is Predictive analytics
                         today?


                  An essay presented to the



    ...
Plagiarism Declaration

1.          I know that plagiarism is wrong. Plagiarism is to use another’s work and pretend that ...
Table of Contents

ABSTRACT..................................................................................................
Anúncio
Anúncio
Carregando em…3
×

Confira estes a seguir

1 de 13 Anúncio

How relevant is Predictive Analytics relevant today?

Baixar para ler offline

This is my view on how relevant is Predictive Analytics relevant today. Although its a high level view, it gives great insights to a person who is looking for somewhere to begin. This was an essay for the

This is my view on how relevant is Predictive Analytics relevant today. Although its a high level view, it gives great insights to a person who is looking for somewhere to begin. This was an essay for the

Anúncio
Anúncio

Mais Conteúdo rRelacionado

Diapositivos para si (20)

Anúncio

Semelhante a How relevant is Predictive Analytics relevant today? (20)

Mais recentes (20)

Anúncio

How relevant is Predictive Analytics relevant today?

  1. 1. How relevant is Predictive analytics today? An essay presented to the Department of Information Systems University of Cape Town By Mugerwa Steven (MGR******) in partial fulfilment of the requirements for the Information and Communication Technologies (INF2010S) 2012 14 September 2012
  2. 2. Plagiarism Declaration 1. I know that plagiarism is wrong. Plagiarism is to use another’s work and pretend that it is one’s own. 2. I have used the APA convention for citation and referencing. Each contribution to, and quotation in, this essay from the work(s) of other people has been attributed, and has been cited and referenced. 3. This essay is my own work. 4. I have not allowed, and will not allow, anyone to copy my work with the intention of passing it off as his or her own work. 5. I acknowledge that copying someone else’s assignment or essay, or part of it, is wrong, and declare that this is my own work. Signature 2 Mugerwa Steven- MGR******
  3. 3. Table of Contents ABSTRACT.................................................................................................................................... 4 INTRODUCTION ........................................................................................................................... 4 I. BACKGROUND ....................................................................................................................................... 4 II. PURPOSE............................................................................................................................................. 4 1. WHAT IS PREDICTIVE ANALYTICS?.......................................................................................... 5 I. DEFINITION .......................................................................................................................................... 5 II. HOW DOES PREDICTIVE ANALYTICS WORK?............................................................................................... 5 III. TYPES OF PREDICTIVE ANALYTICS............................................................................................................ 6 IV. TOOLS ............................................................................................................................................... 6 V. BENEFITS OF PREDICTIVE ANALYTICS ........................................................................................................ 7 2. WHAT ARE THE VARIOUS APPLICATIONS OF PREDICTIVE ANALYTICS?..................................... 7 I. BUSINESS APPLICATIONS ......................................................................................................................... 7 II. FINANCIAL INSTITUTIONS ....................................................................................................................... 8 III. FRAUD AND THREAT ............................................................................................................................. 8 IV. OTHER FIELDS ..................................................................................................................................... 9 3. CHALLENGES AND OPPORTUNITIES INVOLVED WITH PREDICTIVE ANALYTICS.......................... 9 I. CHALLENGES ......................................................................................................................................... 9 II. OPPORTUNITIES .................................................................................................................................10 4. CONCLUSION....................................................................................................................... 11 BIBLIOGRAPHY........................................................................................................................... 12 3 Mugerwa Steven- MGR******
  4. 4. Abstract Predictive analytics can be thought of as analytics of the future. It has a common definition, numerous approaches but has not been exploited to full potential. According to the Gartner Hype Cycles, Predictive analytics is said to achieve its full potential in the next two year. (Gartner, 2012) This paper argues that real-world applications should adopt Predictive analytics in their day to day process in order to stay relevant, productive and ahead of the competition (in profit making firms). The paper goes on to draw an analogy between predictive models and data management and discusses how organizational management can leverage this in order to predict the future and make informed decisions based on those predictions. Introduction I. Background The 21st century is very reliant to information technology and is no wonder it’s known to many as the information age. For our continuous existence, data is by far the World’s most valuable asset. However, data has many forms i.e. data can be raw of which not much can be understood from it and therefore concise decisions won’t always be made. Data is most valuable to us in a processed state normally referred to as information which we can make decisions based on it. In order for data to be able to help us in precise and smart decision making, it has to go through critical analysis known as “analytics”. Analytics is the use of data, statistical and quantitative methods and predictive models to allow organizations and individuals to gain insights into and act on complex issues. Analytics comprises of various forms today e.g. Big Data, Business Intelligence as well as Predictive analytics which will be the basis of this essay. II. Purpose Predictive analytics is the topic of question because it comprises modern phenomenon in practice today such as machine learning (an element of artificial intelligence) as well as the use of past and present data to help in forecasting/predicting the future. The ability to predict the future through predictive analytics explains how valuable data is. More organizations across several industries are using Predictive Analytics as it is a transformational technology that enables more proactive decision making, driving new forms of competitive advantage Also because analytics and business intelligence is ranked number 1 in the technology priorities according to the Gartner EXP Worldwide Survey of 2,300 CIOs - Jan 2012 for increasing 4 Mugerwa Steven- MGR******
  5. 5. enterprise growth. Predictive analytics which is a big part of analytics and business analytics naturally therefore becomes a business priority. Predictive analytics can also support plenty of other business priorities such as growth, productivity etc. Business Intelligence has been regarded a top application and technological development from 2003-2011 (Luftman & Ben-Zvi, 2011) therefore encouraging more entities to adopt Predictive analytics. This essay is setting out to go in detail and explain what predictive analytics is, how predictive analytics can be applied in various disciplines today, how it works, its opportunities and challenges as well as its place in the current technological World. 1. What is Predictive Analytics? I. Definition Predictive Analytics is a branch of business intelligence that uses data mining and statistics to make predictions on future happenings. (Ganesh, Reddy, Manikandran, & Krishna, 2011) Predictive analytics is the branch of data mining (Predictive Analytics is today often referred as data mining) concerned with forecasting probabilities. It is the use of a combination of machine learning, statistical analysis, modeling techniques, and database technology, to process data and uses it to predict future trends and behavioural patterns therefore uncovering problems and opportunities in an organization. These techniques are applied to many disciplines, including marketing, healthcare, financial field like insurance, fraud which will be discussed in more detail. These are usually disciplines in which there's an abundance of data and a need to forecast the future. Predictive analytics helps organizations predict with confidence what will happen next so that smarter decisions can be made and improve objective outcomes. II. How does Predictive Analytics work? Predictive analytics include statistical models and other empirical methods that are aimed at creating empirical predictions (Shmueli & Koppius , 2011) There are many different algorithms used in Predictive Analytics to try to classify patterns, trends and behaviours for a particular variable e.g. for customers. Various models are created in order for Predictive analytics to be possible. These include:  machine learning,  statistical analysis 5 Mugerwa Steven- MGR******
  6. 6.  A combination of various input models using different perspectives (known an ensemble model or a Meta model). Predictive models are not perfect, but they are a lot better than just guessing. For example, if we know that the conversion rate for a promotion is just 3%, it would help to have a good idea of who those 3% of people are so that we can focus on them first. The specific algorithm chosen depends on a combination of the intended use of the prediction e.g. do we need to know why a customer has a certain rank? As well as on how well the algorithm interacts with the data. No algorithm works best with all data in in all situations. What most of the algorithms have in common is how the data is presented to create a predictive investigation whose outcomes can be modelled. Some example algorithms to look at are Logistic Regression, Visualisation and Neural Networks etc. for situations where the behaviour is yes/no. III. Types of Predictive Analytics  Descriptive models It is the task of providing a representation of the knowledge discovered without necessarily modelling a specific outcome. This will be used to categorize or group behaviour in data sets to describe a pattern but nothing beyond that.  Predictive models : However, descriptive analytics is simply not enough. In the society we live in today, it is imperative that decisions be highly accurate and repeatable. For this, organisations are using predictive analytics to literally tap into the future and, in doing so, define sound business decisions and processes. While descriptive analytics lets us know what happened in the past, predictive analytics focuses on what will happen next. IV. Tools Historically Predictive analytics required a specified skill set to do what it does today. But the introduction of Predictive IT analytics systems like Hewlett-Packard’s Service Health Analyzer, IBM’s SPSSpowered Tivoli product, Netuitive’s eponymous offering and other systems make this job much simpler, easier and achieve results quicker. 6 Mugerwa Steven- MGR******
  7. 7. V. Benefits of Predictive Analytics The biggest contribution Predictive analytics gives the World is the fact that it can be used in various industries because of the fact that it works with data to predict the future. Below is a list of how organizations can benefit from the use of Predictive analytics.  It helps to manage performance & risk. It can predict issues prior to and solve any problems such as an outage, degradation in service, or other impacts on business plans  It helps organizations in advanced planning & scheduling capabilities leveraging analytics such as capacity planning, capacity management and workload scheduling  It helps in business optimization. This means a business can constantly adapt to change within dynamic infrastructures  It captures meaningful business insights from operational & business data  It helps identify new business opportunities for profitable growth  Leveraging service and infrastructure analytics, organizations can optimize operations and ensure predictable business outcomes. All in all predictive analytics will be at the forefront to help organizations control costs and acquire a competitive advantage in their industries. 2. What are the various applications of Predictive Analytics? Analytics and predictive analytics will be applied across many domains from banking, insurance, retail, telecom, energy etc. The existence of various analytical software as well as high levelled skill sets make Predictive analytics possible. Predictive analytics can be applied to more than one industry simply because of its ability to generate useful predictions that companies can use to make informed decisions. Predictive analytics uses statistical analysis and predictive modelling in order to make proactive decisions. This means that entities make decisions prior which is preferred to reactive decision making which is merely a response to a setback or a change in business operations. Below are the various ways in which Predictive analytics is applied in the real World. I. Business Applications Predictive analytics is revolutionizing the way companies do business today. The greatest benefit of deployment for any predictive system is reaped when predictive analytics is integrated into business processes. The most commonly used applications of Predictive analytics in business are Enterprise Resource Planning (ERP) and Customer Relations Management (CRM) applications. 7 Mugerwa Steven- MGR******
  8. 8. ERP consists of resource management for a particular business. Businesses use predictive analytics in supply chain management to manage stock levels (just-in-time). Revenues can also be forecasted by looking into past sales data and use a time series analysis. Organizations can predict the next point or two forward in a series, and then as more real data is gathered, predictions are made. Customer relationship management (CRM) systems perform the tasks of monitoring activities, coordinating resources, and generally keeping your organization on track with its sales processes. In business, predictive analytics are often used to answer questions about customer behaviour. For example, companies often want to know whether or not a particular customer is likely to be interested in a particular offer or whether a new customer will become a long-term customer given a certain set of premiums and benefits. Therefore predictive analytics helps business to segment their customers into understandable groupings as well as calculate metricises such as reorder rates, seasonality by customer type, targeted marketing, and selling initiatives. This will therefore make marketing strategies much simpler and cost effective as an organisation now has information about particular customers. Ultimately, businesses want predictive analytics to suggest how to best target resources for maximum return. This way it uncovers hidden insights from data so one can create personalized experiences that will reduce business costs, increase customer loyalty and also identify risks that could derail entity plans and take timely corrective action (proactive decisions over reactive). II. Financial Institutions Financial institutions have been able to adopt the use of predictive analytics very smoothly into their infrastructure. Predictive analytics is used by banks, micro-finance, retailers and insurers to calculate credit scores. Predictive analytics is used to calculate organisation and individuals credit scoring. A credit score is a figure processed through tracking of a customer’s credit history, loan application, earnings in order to predict future creditworthiness of individuals/entities. Lenders i.e. banks, micro-finance and other specialists use Predictive analytics to determine who qualifies for a loan as well as which customers will bring in the most revenue. Credit scoring is used throughout the credit industry in South Africa. III. Fraud and threat This is mainly used by Insurance companies and to an extent banks. South African firms have been able to use Predictive analytics to monitor their business environment, detect suspicious activity, and control outcomes to minimize loss. 8 Mugerwa Steven- MGR******
  9. 9. By using IBM SPSS predictive analytics to identify risks and accelerate claims settlement, Santam Insurance boosted customer service and managed to beat fraud. "In the first month of using the SPSS solution, we were able to identify patterns that enabled us to foil a major motor insurance fraud syndicate. Within the first four months, we had saved R17 million on fraudulent claims, and R32 million in total repudiations – so the solution delivered a full return on investment almost instantly!" - Anesh Govender, Head of Finance, Reporting and Salvage, Santam Insurance (IBM, 2011) IV. Other fields  Predictive analytics is used health care to determine which patients are at risk of developing particular conditions.  Predicting crime  Predictive analytics is already being used in traffic management in identifying and preventing traffic gridlocks.  Operational activities to ensure staff, processes and assets are aligned and optimized to maximize productivity and profitability.  Applications have also been identified for energy grids, for water management.  Risk Management  Educational institutes to predict student grades. 3. Challenges and Opportunities involved with Predictive Analytics I. Challenges It is not always easy to incorporate Predictive analytics in any organisation due to various challenges faced in the workplace. This could consist of both internal and external constraints of an organization making it a struggle for organizations to find a balance during implementation. These challenges are compiled in the table below. Challenge Description Technical Factors  Data Quality; the aspect of data is very important as it is the core ingredient for predictive analytics to work. This means data has to be consistent, readable and accurate. Data also needs to be stored securely.  System Architecture; this entails the current systems in place at a particular workplace or organization. The 9 Mugerwa Steven- MGR******
  10. 10. software must be in sync with other systems in place or risk disrupting business operations.  Resources; this involves the level of infrastructure i.e. hardware, networks etc. to support predictive analytics.  Team Skills; this is by far another important aspect as without professionals, data is of no use to the organization. Organisational and  Business Focus; this is the business vision and policies Management Factors that it follows to attain its objectives. Some organisations are not entirely in need of Predictive analytics even with the information it offers individuals.  Company politics and Management Support; this is important as management depicts the business direction. Thus if it adopt Predictive analytics with a positive view it will definitely succeed. However, management support in most corporations is sluggish on adoption of new technologies and therefore leads to a challenge. User Participation  Commitment; A resistance to change is usually experienced by workers in a workplace who don’t want to undergo training and use new technologies.  Project Management is difficult as communication about new technologies is never easy. These issues in a sense therefore also depict variables that need to be in place for Predictive analytics to be a success. II. Opportunities There is absolutely no question that predictive analytics will be pervasive across a wide range of applications. It will be everywhere. Integrations with other technologies such as big data and cloud computing. 10 Mugerwa Steven- MGR******
  11. 11. Big Data is a term used to describe large and complicated data sets that can’t be worked on using traditional database management. The big question pertaining to Big Data are "how to extract insights and value from it as well as being effective about it". The answer is predictive analytics. Cloud Computing is a set of services that provides computing resources via the Internet. Large data centers deliver scalable, on-demand resources as a service, eliminating the need for investments in specific hardware or software, or on organizational data center infrastructure. It allows for a variety of services, including storage capacity, processing power, and business applications. With the power of Predictive analytics and technologies like cloud computing, big and small organizations could save millions, be more productive and efficient at the same time. Therefore, Predictive analytics function is not limited to what it can do, but also to what it can achieve once it is associated with other technologies in an infrastructure. 4. Conclusion This paper shows my views on how predictive analytics influences the world today as well as the step process involved in making Predictive analytics possible. The world is heavily reliant on technologies and the ease brought forward by various tools doesn’t make Predictive analytics an exception. Although still not widely used in the world, Predictive analytics has massive potential to change the way we think and leave our lives. It definitely has the potential to grow rapidly over the following years in order to make predictions and most importantly stays relevant to our societies. 11 Mugerwa Steven- MGR******
  12. 12. Bibliography Apte, C. V., Hong, S. J., Natarajan, R. R., Pednault, E. D., Tipu, F. A., & Weiss, S. M. (2003). Data- intensive analytics for predictive modeling. IBM Journal Of Research & Development, 47(1), 17. Baecke, P., & Van Den Poel, D. (2010). IMPROVING PURCHASING BEHAVIOR PREDICTIONS BY DATA AUGMENTATION WITH SITUATIONAL VARIABLES. International Journal Of Information Technology & Decision Making, 9(6), 853-872. doi:10.1142/S0219622010004135 Bradley, P. (2012). Predictive analytics can support the ACO model. Hfm (Healthcare Financial Management), 66(4), 102-106. DAVENPORT, T. H., & HARRIS, J. G. (2009). What People Want (and How to Predict It). MIT Sloan Management Review, 50(2), 23-31. Ganesh, M. S., Reddy, C. P., Manikandran, N., & Krishna, P. V. (2011). TDPA: Trend Detection and Predictive Analytics. International Journal on Computer Science & Engineering, 3(3), 1033-1039. Gartner. (2012, August 16). Press Resources: Gartner. Retrieved September 14, 2012, from Gartner Web Site: http://www.gartner.com/it/page.jsp?id=2124315 Hair, J. F. (2007). Knowledge creation in marketing: the role of predictive analytics. European, 19(4), 303-315. IBM. (2011, July). Case Studies:International Business Machines. Retrieved September 13, 2012, from An International Business Machines Web site: http://www- 01.ibm.com/software/success/cssdb.nsf/CS/STRD- 8JJETD?OpenDocument&Site=default&cty=en_us Luftman, J., & Ben-Zvi, T. (2011). Key Issues for IT Executives. MIS Quarterly Executive, 10(4), 203-212. 12 Mugerwa Steven- MGR******
  13. 13. Shmueli, G., & Koppius , O. (2011). PREDICTIVE ANALYTICS IN INFORMATION SYSTEMS RESEARCH. MIS Quarterly, 35(3), 553-572. 13 Mugerwa Steven- MGR******

×