July 22nd 2016

Wise Practitioner – Predictive Analytics Interview Series: Dae Park and Vijay D’Souza at Government Accountability Office (GAO)

By: Sean Robinson, Program Chair, Predictive Analytics World for Government

In anticipation of their upcoming conference co-presentation, Characteristics for Those Dae Park IMAGEClaiming Social Security Benefits Early, at Predictive Analytics World for Government, October 17-20, 2016, we asked Dae Park, Assistant Director at Government Accountability Office (GAO), and Vijay D’Souza, Director at Government Accountability Office (GAO), a few questions about their work in predictive analytics.

Q:  How would you characterize your agency's current and/or planned use of predictive analytics?  What is one specific way in which predictive Vijay D'Souza IMAGEanalytics actively drives decisions in your agency?

A: A core function of the Government Accountability Office (GAO) is to use data to objectively evaluate government programs and spending. We both develop our own models and evaluate agencies’ own models to evaluate the effectiveness of government programs. For example, we have completed numerous evaluations of Medicare payment policies to evaluate possible effects on quality and access to care.

Q:  Can you describe the challenges you face or have already overcome in establishing a data-driven environment in your agency?

A: The biggest problem we face is acquiring and validating data from the agencies we audit. Data are provided in multiple formats and often have quality issues. These include duplicate, incomplete, or inaccurate entries, and inconsistencies among records. This challenge is compounded when we merge data from multiple agencies or multiple sources to conduct more sophisticated analyses.

Q:  Can you discuss any near term goals you have for improving your agency's use of predictive analytics?

A: There is an increasing awareness across our analyst community of the value of predictive analytics to measure the effectiveness of federal programs. This includes training and outreach efforts from our analytics experts to other agency staff. We are also taking greater advantage of open source tools for analytics and developing strategies to increase the amount of staff that can help with the initial stages of data acquisition, evaluation, and preparation.

Q:  Can you describe a successful result from the employment of predictive analytics in your agency, i.e., cost avoidance, funds recovered, improved efficiency, etc.

A: We analyzed TSA’s program to detect more risky air travel passengers and determined it was not based on sound evidence. In another case, we evaluated a Department of Transportation program that used a large amount of data to calculate safety scores for commercial trucking carriers and to ostensibly determine the likelihood of crashes. We found that violations used in the agency model either occurred too infrequently to be useful, or were not otherwise predictive of crashes.

Q:  Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World for Government.

A: To better understand the circumstances faced by those who claim early Social Security benefits, GAO examined demographic and occupational characteristics associated with early claiming. More specifically, we examined the characteristics and income of early claimers using data from the Health and Retirement Study. I look forward to discussing our findings in detail at the session.

———————

Don't miss Dae and Vijay’s conference co-presentation, Characteristics for Those  Claiming Social Security Benefits Early, on Monday, October 17, 2016 from 11:25 am to 12:10 pm at Predictive Analytics World for Government. Click here to register to attend.

By: Sean Robinson, Program Chair, Predictive Analytics World for Government

No Comments yet »

July 19th 2016

Wise Practitioner – Predictive Analytics Interview Series: Dean Abbott of SmarterHQ

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, The Revolution in Retail Customer Intelligence, at Predictive Analytics World New York, October 23-27, 2016, we asked Dean Abbott, Co-Founder and Chief Data Scientist of SmarterHQ, a few questions Dean Abbott imageabout his work in predictive analytics.

Q: In your work with predictive analytics, what behavior do your models predict?

A: I’ve built models that predict a wide variety of behaviors and patterns. A short list is provided here:

  • Customer behavior: response, churn, product up-sell and cross-sell, best marketing creative, days to next purchase, days to next visit;
  • Signals (radar): tank, truck column of tanks; (sonar): man-made vs. biologic;
  • Financial: fraud or suspicion of fraud, debt repayment period, debt repayment amount, insurance claim repayment likelihood, claim amount of repayment.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A:  I’ll speak to SmarterHQ, the company I’m Co-Founder of and Chief Data Scientist. There is no one specific way our predictive models drive decisions, but they are involved in the decision-making process in several ways, all related to selecting customers to promote to, whether that be selecting customers to send an email to, show a display ad, or present content on a page that is of greater interest to the customer.

Consider our models that predict the likelihood that someone will purchase a product during a visit to the company’s web site within 3 days. Each visitor is scored while they browse on the website and at the end of their session. The company now wants to create a new campaign to increase sales of a particular product by emailing them a promotion code with a 20% discount. If the customer is likely to purchase a product on the web site within 3 days, the models will exclude these customers from the email list; why take away margin from sales that are likely to occur anyway. Or what if a customer was very likely to purchase within 7 days last week but is no longer likely this week? This is a form of churn (but based on expected behavior, not actual behavior), and these customers could be given incentives to visit again.

Q: Can you describe a successful result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: It is difficult to describe most of the results my models have generated because they are considered sensitive information for the company or government agency. I’ve had models in use by organizations for 10 years before they were refreshed. I’ve had another model so successful that it was put on the “do not tell” list by the organization because it became a strategic initiative for the organization. I’ve had fraud models identify multi-million dollar cases to investigate that were clearly fraud but had previously eluded detection.

Q: What surprising discovery have you unearthed in your data?

A: Most surprising? There have been many surprises over the years, usually related to the data itself and patterns of behavior that we may overlook, but are important nevertheless. For example, with the days to next purchase models, one expects that visitors on a website who look at lots of hot products are more likely to purchase soon; these are engaged visitors. However, it turns out that some of the most likely purchasers are those who visit just one item. The vast majority of the time, one-item visitors are not engaged and therefore are unlikely to purchase. But, if these one-item visitors were previously highly engaged, it’s a different story; they are focused like a laser beam on one product only. So the surprise was that there is this subset of visitors who look awful but are actually fantastic!

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: The most important take-away in my talk is this: When you prepare data for modeling, think about how the algorithms interpret the data. Each algorithm has weaknesses that can result in strange or misleading behavior. It’s our job as predictive modelers to help the algorithms do the best job they can.

Q: In addition to keynoting, you will be teaching two one-day workshops at PAW New York, Supercharging Prediction with Ensemble Models and Advanced Methods Hands-on: Predictive Modeling Techniques. How would you advise attendees to choose between these workshops and would it even make sense to attend both?

A: There are many workshop options, and all of them are worthy of attending. I think of the Supercharging and Advanced Methods workshops as complementary to your Online Introduction to predictive and John Elder’s Modeling Methods, with the sequence being (1) Intro, (2) Modeling Methods, (3) Advanced Methods, and (4) Supercharging. The Modeling Methods can be taken the day before Advanced Methods in the same conference; Modeling Methods provides a framework for predictive modeling, and Advanced Methods lets you try it out on commercial software. Supercharging takes predictive modeling to the next level, introducing the methods that win modeling competitions and have provided me with extra accuracy has made the difference between successful models and very successful models in my consulting practice.

———————

Don't miss Dean Abbott’s conference presentation, The Revolution in Retail Customer Intelligence, and workshops at Predictive Analytics World New York on Tuesday, October 25, 2016 from 10:30-11:15 amClick here to register for attendance.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »

June 16th 2016

Wise Practitioner – Manufacturing Predictive Analytics Interview Series: Chris Labbe at Seagate Technology

By: Bala Deshpande, Conference Co-Chair, Predictive Analytics World for Manufacturing 2016

In anticipation of his upcoming Predictive Analytics World for Manufacturing conference presentation, Building a Predictive Analytics Organization, we interviewed Chris Labbe, Chris Labbe HeadshotManaging Technologist at Seagate Technology. View the Q-and-A below for a glimpse of what’s in store at the PAW Manufacturing conference.

Q: What are the challenges in translating the lessons of predictive analytics from other verticals into manufacturing?

A: ​Manufacturing at a company like Seagate means volume.  Volume of parts produced is 10’s of millions per quarter and each of these devices generates many MB of data.  Daily we produce several TB of data from the drives themselves.  Then we have vertically integrated components generating a couple more TB per day and a massive number of machines that will benefit from advanced sensors.  This magnitude and velocity of data is well beyond the typical retail / marketing analytics challenge and stresses our IT systems to the limits.

Q: In your work with predictive analytics, what behavior do your models predict?

A: Although we built this team with a desire to help fix engineering problems, we found much stronger partners in the business side of the company.  Partially this is because engineering is pretty structured in data management and resistant to “advanced analytics” since they often feel they are using advanced techniques already.  Meanwhile the business teams know that they need help.  As such, not only are we working on better methods for quality management of the production system, but also ideas like customer ordering predictability, supply chain management, inventory reduction and build-ahead risk.  In a way, these are all manufacturing challenges since bad decisions in the business front lead to inefficiencies in production.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A: Our business is pretty mature, which means we have a lot of people that have been using data effectively yet inefficiently.  Some of the most important projects we are engaging in are attempts to pull the company into machine learned multi-variate methods instead of human biased univariate system for quality control.​

Q: Is your company supportive of the work your team is doing?  And are they well prepared to execute on the models and systems you develop?

A: We are very fortunate to have the attention of the President of Engineering, Manufacturing & Sales.  This means a lot in terms of stability for the team while we develop some of the projects as it can take a while to demonstrate effectivity.  Meanwhile, we came into this effort fairly unprepared for how big the gaps are in Seagate’s data management pipeline.  To move Seagate to a World Class Data Science company is going to take a lot of time and a lot of money.​

Q: What surprising discovery have you unearthed in your data?

A: There are several projects that we have been pulled in to help with visualizations and data automation.  What we see behind the scenes is often pretty scary, though.  From manual data processes that push sensitive data through email, to data manipulation between source and decision and even weak statistical methods being applied to the data.  Turns out we can help the company in many more ways than just machine learning.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World for Manufacturing.

A: In building this new division at Seagate, there have been many lessons to learn.  Probably the most important of them all is how to invest efforts up front on data engineering and BI visualization tools.  This has given us the “keys to the castle” by allowing the team to fully understand the underlying math in a tool.  Once the target group is excited about the improvements in the tool, then we can begin a discussion about improving the model behind the scenes.​  When we have started a project with “we can make your model better,” the progress is slow at best.​

————————

Don't miss Chris’ conference presentation, Building a Predictive Analytics Organization, at PAW Manufacturing, on Tuesday, June 21, 2016, from 4:45 to 5:30 pm. Click here to register for attendance. 

By: Bala Deshpande, Founder, Simafore and Conference Co-Chair of Predictive Analytics World for Manufacturing.

No Comments yet »

June 13th 2016

Wise Practitioner – Manufacturing Predictive Analytics Interview Series: Peter Frankwicz at Elmet Technologies

In anticipation of his upcoming Predictive Analytics World for Manufacturing conference presentation, Improved Statistical Process Control of Mature Manufacturing Processes Using Multiple Available Data Streams, we interviewed Peter Frankwicz, Senior Process Engineer at Elmet Peter Frankwicz imageTechnologies. View the Q-and-A below for a glimpse of what’s in store at the PAW Manufacturing conference.

Q: What are the challenges in translating the lessons of predictive analytics from other verticals into manufacturing?

A: Most small business manufacturing companies are focused on relatively simple statistical process control (SPC) or end-of-line quality control.  The next step to predictive statistical process control is a major undertaking in both collection of relevant manufacturing process data and product & yield metrics.  The data-based “return on investment” in higher yield has to overcome management angst of higher risk of product scrap at the end of the manufacturing line.

Q: In your work with predictive analytics, what behavior do your models predict?

A: Predictive statistical process control analytics are in use to optimize powder metallurgical properties, such as tap density, and sintered ingot properties for thermomechanical processing to sheet and rod products.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A: Many small business manufacturing companies have a “we have always made it this way” mentality.  Predictive analytics and statistical process control methods allows process engineering to deliver data-based and statistically significant understanding of manufacturing processes to management.  Predictive analytics drives several specialized [Elmet Technologies refractory metal product – manufacturing process] combinations to optimize yield and quality.

Q: Can you describe a successful result, such as the predictive lift (or accuracy) of your model or the ROI of an analytics initiative?

A: Predictive models were developed to understand refractory metal powder properties.  Use of these models to direct manufacturing production resulted in an over $20,000 monthly reduction in scrap product in the “downstream” sheet rolling department.

Q: What surprising discovery have you unearthed in your data?

A: Data mining and statistical analysis of available process data revealed surprising manufacturing process sensitives.  Many of these process sensitives were only known at the level of tribal knowledge on the manufacturing floor.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World for Manufacturing.

A: There is a high need in small business manufacturing for simple and robust predictive analytical methods.  Even starting the path with data mining and statistical analysis of available data streams can discover surprising and valuable manufacturing process insights and yield optimization strategies.

————————

Don’t miss Peter’s conference presentation, Improved Statistical Process Control of Mature Manufacturing Processes Using Multiple Available Data Streams , at PAW Manufacturing, on Tuesday, June 21, 2016, from 2:40 to 3:25 pm. Click here to register for attendance.

By: Bala Deshpande, Founder, Simafore and Conference Co-Chair of Predictive Analytics World for Manufacturing.

 

 

No Comments yet »

June 10th 2016

Wise Practitioner – Text Analytics Interview Series: Dirk Van Hyfte at InterSystems Corporation

By: Steven Ramirez, Conference Co-Chair of Text Analytics World Chicago

In anticipation of his upcoming conference co-presentation, Personalized Medicine and Text Analytics at Text Analytics World Chicago, June 21-22, 2016, we asked Dirk Van Dirk Van Hyfte imageHyfte, Senior Advisor for Biomedical Informatics at InterSystems Corporation, a few questions about his work in text analytics.

Q: In your work with text analytics, what behavior or outcome do your models predict?

A: To support the shift from reactive to pro-active medicine we look for patients who are at risk to develop Sepsis, Hepatitis C and Delirium. In the area of Behavioral Health we support harm reduction projects.

Q: How does text analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: The ability to effectively harness the mountains of unstructured data in healthcare is a key strategic asset for any successful organization.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: One California IDN has demonstrated great success in reducing sepsis mortality, bringing it down from 55% at the start of their interventions to a current level of 35%.  In an effort to bring sepsis mortality down even further to 25%, this IDN is now piloting the use of Text Analytic Tools to add unstructured data analysis to its armamentarium.

Q: What surprising discovery or insight have you unearthed in your data?

A: There are huge gaps in structured data fields. At a Cancer Registry we identified a data gap of 22% where HER-2 testing had been ordered but no definite outcome was recorded. The organization was able to identify a systematic shortfall in the availability of results and was able to investigate this.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Text Analytics World.

A: Numerous case studies, a few of which have been discussed above, have demonstrated the shortcomings of structured data.  Important information, such as disease risk factors, might not exist within structured data at all, and even when the appropriate structured data field exists within a data model, the data can be missing or unavailable.  Unstructured data, which comprises perhaps 80% of all healthcare data, has great potential to replace that missing structured data and/or complement what’s already there. 

———————

Don't miss Dirk’s conference co-presentation, Personalized Medicine and Text Analytics on Wednesday, June 22, 2016 from 3:30 to 4:00 pm at Text Analytics World Chicago. Click here to register to attend.

By: Steven Ramirez, CEO at Beyond the Arc, and Co-Chair of Text Analytics World

No Comments yet »

June 6th 2016

Wise Practitioner – Text Analytics Interview Series: Michael Dessauer and Justin Kauhl at The Dow Chemical Company

By: Steven Ramirez, Conference Co-Chair, Text Analytics World Chicago

In anticipation of their upcoming conference co-presentation, Understanding our Customers' Customers' Customers' Needs – Text Analytics for B-to-B Businesses at Text Michael DessauerAnalytics World Chicago, June 21-22, 2016, we asked Michael Dessauer, Data Scientist at The Dow Chemical Company and Justin Kauhl, Computational Linguistics Expert at The Dow Chemical Company, a few questions about their work in text analytics.

Q: In your work with text analytics, what behavior or outcome do your models predict?

A:  Our team, Dow’s Advanced Analytics Group, service company-wide needs from any Justin Kauhlof our businesses or functions so our project outcomes are very diverse. In general, we get many requests for market insight which boils down to consumer needs, so we try to categorize product features and sentiment around those features.

Q: How does text analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: The delivered value statements have come from businesses using our market listening insights to secure value pricing on new product introductions. It’s easier to sell when we can point to overwhelming consumer needs that our analysis uncovers.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: In one case Dow was able to value price a product by 10X and our insights were given credit for contributing to that increase.

Q: What surprising discovery or insight have you unearthed in your data?

A: We have had success using external unstructured data to improve short-term forecasting. It was a bit unexpected but very helpful for some of our existing modeling efforts.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Text Analytics World.

A: We intend to share our experiences, both successful and unsuccessful, in developing effective text analytics projects with internal clients. We will share our process, technologies employed, and our best practices learned over the past several years working on text analytics projects.

———————

Don't miss Michael and Justin’s conference co-presentation, Understanding our Customers' Customers' Customers' Needs – Text Analytics for B-to-B Businesses on Tuesday, June 21, 2016, from 1:30 to 2:15 pm, at Text Analytics World Chicago. Click here to register to attend.

By: Steven Ramirez, CEO at Beyond the Arc, and Co-Chair of Text Analytics World

No Comments yet »

June 3rd 2016

Wise Practitioner – Manufacturing Predictive Analytics Interview Series: Edward Crowley at The Photizo Group, Inc.

By: Bala Deshpande, Conference Co-Chair, Predictive Analytics World for Manufacturing 2016

In anticipation of his upcoming Predictive Analytics World for Manufacturing conference Edward Crowley imagepresentation, Predictive Analytics – What is 2% Worth, we interviewed Edward Crowley, CEO at The Photizo Group, Inc. View the Q-and-A below for a glimpse of what’s in store at the PAW Manufacturing conference.

Q: What are the challenges in translating the lessons of predictive analytics from other verticals into manufacturing?

A: The biggest challenge in any predictive analytics deployment is to have the understanding of the business processes, business models, and workflows required to build predictive analytics models that have a meaningful impact on the business. Clearly, these processes and models can be unique to each vertical. The bigger issue is, in my mind, that much of predictive analytics has been focused on marketing and customer facing applications; however, there are tremendous opportunities for business process and operational related predictive analytics solutions which bring immediate savings to your organization or your customer’s organization and which are often overlooked when the focus is on customer facing solutions.

Q: In your work with predictive analytics, what behavior do your models predict?

A: Today our models predict device failures for either A) very expensive capital equipment items which have significant costs associated with lost production capacity due to unexpected failures, or B) mass fleets (as high as several million units) of devices which benefit from predicting failure in order to reduce service costs, improve productivity, and early identification of mass failures.

Q: How does predictive analytics deliver value at your organization? What is one specific way in which it actively drives decisions?

A: By understanding when a printer or copier will actually be out of toner (versus when it is ‘low’ on toner), the amount of toner left in the cartridge when it is replaced can be reduced from up to 35% of total capacity to less than 10% of capacity. In this example, our model accurately predicts when toners will be empty, shipping the toner ‘just in time’ before the printer runs out versus shipping the toner when the printer initiates a toner low alert.

Q: Can you describe a successful result, such as the predictive lift (or accuracy) of your model or the ROI of an analytics initiative?

A: Our first model, a toner delivery optimization model, reduces ‘lost toner’ and excess shipping costs by over 50% per device. This translates into over $100 per year, per device savings – or from $5M to $50M for each of our OEM customers. The ROI for the model is immediate since we deliver this in a ‘as a Service’ model where the client just has to sign up; we have already developed the solution.

Q: What surprising discovery have you unearthed in your data?

A: Initially, we thought that the wasted toner from throwing away toner cartridges early was around 15% of the total cartridge capacity – but a validation phase of our project where we actually measured the amount of toner in a large volume of returned cartridges identified that the average waste is closer to 32% per cartridge.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World for Manufacturing.

A: There are, in my view, three key building blocks to PA – industry knowledge, PA technology, and a knowledgeable team which can turn technology into a solution. It’s not just about the software – it’s about the industry knowledge!

————————

Don't miss Edward’s conference presentation, Predictive Analytics – What is 2% Worth, at PAW Manufacturing, on Wednesday, June 22, 2016 from 3:30 to 4:15 pm. Click here to register for attendance. 

By: Bala Deshpande, Founder, Simafore and Conference Co-Chair of Predictive Analytics World for Manufacturing.

No Comments yet »

May 31st 2016

Wise Practitioner – Predictive Analytics Interview Series: Tanay Chowdhury at Zurich North America

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, Deep Learning in Cloud Based Applications at Predictive Analytics World Chicago, June 20-23, 2016, we asked Tanay Tanay Chowdhury imageChowdhury, Associate Data Scientist at Zurich North America, a few questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior or outcome do your models predict?

A: They predict different things from predicting risk to hire a contractor, predict crop yield, to how well a software is performing given its telemetry data.

Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: It helps immensely in building better, customer oriented software’s. The decision to include new feature in an existing piece of software basing on user experience in related features is one such instance.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: My predictive model to predict crop yield in economic perspective helped save the organization approximately 20 million per year compared to the existing model, as the new model was better considering weather variables.

Q: What surprising discovery or insight have you unearthed in your data?

A: Neural network is better in predicting huge fluctuation in output compared to Random Forrest or GBM, and necessarily in such cases ensembling is of little help.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: It will give emphasis on current state of operation on cloud based deep learning implementation challenges and different outlooks related to that.

———————

Don't miss Tanay’s conference presentation, Deep Learning in Cloud Based Applications on Tuesday, June 21, 2016 from 4:20 to 4:40 pm at Predictive Analytics World Chicago. Click here to register to attend.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »

May 23rd 2016

Wise Practitioner – Predictive Analytics Interview Series: Lawrence Cowan at Cicero Group

By: Eric Siegel, Founder, Predictive Analytics World

In anticipation of his upcoming conference presentation, Data Driven Selling: Enabling a Direct Salesforce with Tools that Re-Enforce Predictive Selling Methods at Predictive Analytics World Chicago, June 20-23, 2016, we asked Lawrence Cowan, Partner at Cicero Lawrence_Cowan imageGroup, a few questions about his work in predictive analytics.

Q: In your work with predictive analytics, what behavior or outcome do your models predict?

A: A good portion of the advanced analytics work we do at Cicero deals with consumer behavior across all stages of the customer lifecycle.  So this would span from acquisition through the development of “typing” tools for targeting and segmentation, to response and uplift modeling for existing customers, to attrition modeling and customized intervention strategies.

Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: Predictive analytics defines our organization.  With it, we would not have an offering.  As a full service data-driven strategy consulting firm, it is our job to provide the technical and analytical expertise to help our clients leverage data to make smarter decisions.  And in all engagements involving predictive analytics, our ultimate objectives are results and implementation – if our clients cannot actively use the models and insights to make decisions, we have failed.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: In a recent effort to help a financial institution optimize its marketing campaigns, we developed a model that identified customers who were more likely to respond to a CD campaign, and who were more likely to bring “new money” to the CD (opposed to simply shifting money from another account at the financial institution).  The results were very impressive, including the following metrics: 14x higher response rate, 4x increase in average deposits, 60% in “new money” compared to just 3% “new money” in the control group.

Q: What surprising discovery or insight have you unearthed in your data?

A: I’m always surprised at the opportunities unearthed with predictive analytics – things that you would never expect if it were not for efforts in data mining.  For example, for a large grocery retailer, we were able to identify two critical customer behavior trends (made possible through loyalty data) that were significant predictors of customer profitability.  These two trends were counter to heuristic judgment at the executive level (executives have since changed their perception of the event after seeing the compelling evidence.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.

A: How to leverage secondary data (household level data) to drive more business value from your predictive models.

———————

Don't miss Lawrence’s conference presentation, Data Driven Selling: Enabling a Direct Salesforce with Tools that Re-Enforce Predictive Selling Methods on Tuesday, June 21, 2016 from 10:30 to 11:15 am at Predictive Analytics World Chicago. Click here to register to attend.

By: Eric Siegel, Founder, Predictive Analytics World

No Comments yet »

May 20th 2016

Wise Practitioner – Text Analytics Interview Series: Pengchu Zhang and John Herzer at Sandia National Laboratories

By: Steven Ramirez, Conference Co-Chair Text Analytics World Chicago

In anticipation of their upcoming conference co-presentation, Enhancing search results relevance using Word2Vec Language Models at Text Analytics World Chicago, June 21-22, 2016, we asked Pengchu Zhang, Computer Scientist at Sandia National Laboratories, and Pengchu Zhang imageJohn Herzer, Enterprise Search Project Lead at Sandia National Laboratories, a few questions about their work in text analytics.

Q: In your work with text analytics, what behavior or outcome do your models predict?

A: We use the Word2Vec Neural Network model in our search application to predict word John Herzer imageusage in our corpus for a particular context.  Word2Vec consists of two models, the Continuous Bag of Words (CBOW) model and the Skip-Gram model.  The CBOW model lets us predict a target word given the surrounding words and conversely, the Skip-Gram model lets us predict the surrounding words given a specific word.  We use this capability to enhance our queries with term expansion.

Q: How does text analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?

A: We’re able to increase the relevant content that we return in our search results by enhancing the customer’s query with related terms or synonyms.  This automated way of identifying synonym-like terms is much more cost effective than trying to build and maintain a corporate synonym dictionary.

Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?

A: For one particular acronym query, ‘SAR’, our use of Word2Vec to expand the query into its definition of ‘synthetic aperture radar’ resulted in a 432% increase in relevant documents returned by the search engine.

Q: What surprising discovery or insight have you unearthed in your data?

A: We discovered that there were many more references to “buckyballs”, the spherical carbon molecule in our corpus than we realized.  Use of the Word2Vec model resulted in our query being expanded to include the word ‘fullerenes’, a term more commonly used in scientific papers for this molecule.

Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Text Analytics World.

A: Using the Word2Vec model for query enhancement can help your enterprise move towards conceptual search.

———————

Don't miss Pengchu and John’s conference co-presentation, Enhancing search results relevance using Word2Vec Language Models on Tuesday, June 21, 2016 from 3:35 to 4:20 pm, at Text Analytics World Chicago. Click here to register to attend

By: Steven Ramirez, CEO at Beyond the Arc, and Co-Chair of Text Analytics World

No Comments yet »

Next »

  • Subscribe Via Email

    Get a daily digest of new posts delivered to your inbox:

  • Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die
    BOOK AWARDS:
  • Recent Posts