Select a topic:
All
Data Science
Operations and Supply Chain
  • image of bank
    On Dec 11, the Wall Street Journal’s lead article in its Money and Investing Section was on the news that the Bank of America (BofA) passed its stress test.  This was big news because earlier in the year, the Fed had decided that BofA’s stress test models did not meet standards.  BofA had to spend... Read More »
  • image of bank
    Fallout from the 2008-2009 financial crisis included the emergence of a new regulatory landscape intended to safeguard the U.S. banking system from a systemic collapse. In 2012, the Federal Reserve Board of Governors (Fed), began to require the largest U.S. Bank Holding Companies (BHCs) to file a Comprehensive Capital Analysis and Review (CCAR), with stress tests intended to assess... Read More »
  • Machine Learning and High Quality Potato Chips

    Michael Watson Ph.D Dec 16th, 2015
    iStock_000077072789_Full Potato chips web
    I always like learning about new applications of machine learning algorithms to improve operations.  And, I especially like ones that are easy to explain. I ran across one such application in a podcast interview of Frito-Lay’s Brendan O’Donohoe where he was discussing the potato chip supply chain.  One of the interesting stories he told was... Read More »
  • Introduction Video to Opex Analytics

    Michael Watson Ph.D Dec 14th, 2015
    iStock_000041990766_Large Chicago Skyline for Web
    Be sure to check out our 2 and 1/2 minute introduction to Opex Analytics.
  • Simple Models for Next Day or Next Hour Delivery

    Michael Watson Ph.D Dec 2nd, 2015
    20151122_Winter Path Web Size
    (This article first appeared on SupplyChainDigest) When we wrote the book Supply Chain Network Design, we intentionally started with a very simple model to help people build intuition on how the math works.  This simple model optimized the location of facilities to minimize the distance to the customers.  This model did not consider costs, capacities,... Read More »
  • Podcast: Big Data Analytics with Aster R

    Michael Watson Ph.D Nov 23rd, 2015
    iStock_000046199458_Large_Podcast_Web
    The popularity of R continues to grow.  However, it can be difficult to use R with large data sets. Diego recently recorded a 6-minute podcast on the benefits of Aster R.  In this 6-minute podcast, you will learn about the trade-offs when trying to analyze large data sets with R and how Teradata’s Aster R... Read More »
  • 6 Modeling Tips from LLamasoft and JLL

    Michael Watson Ph.D Nov 16th, 2015
    Man interacting with virtual world map
      (This article first appeared in SC Digest.) At the same CSCMP talk as Benjamin Moore (covered here), Jason Brewer of LLamasoft (a provider of supply chain design software) and Kelly Gray of JLL (or also known as Jones Lang LaSalle, a full services commercial real estate firm) opened up the discussion with some of... Read More »
  • Andy Picture MT 1 Web
      (This article first appeared on SC Digest.  It was a report from the annual CSCMP Conference) With the current low price of diesel fuel and the expected continuing drop in the price of oil, I was wondering how much interest there would be in the CSCMP talk on natural gas trucks.  My question was... Read More »
  • image of bank
    Dodd-Frank Act Stress Testing (DFAST) is now required also for smaller banks with assets less than $50 billion. Simply stated, the test requires a bank to assess possible losses in future years from the loans on the books. A report exploring many different future economic scenarios such as increases in the unemployment rates or short... Read More »
  • iStock_000031281272_Double for Web
    Alex Scott recently published an article on real-time index for truckload rates.  (The article is here but you may need to do a free registration to get access.)   Even though the truckload market is a $700B industry, Alex points out that there is no agreed upon index that shippers and carriers can use to... Read More »
  • Birch Trees small
    In a recent engagement, we were tasked with evaluating sets of time series data at many intervals to determine the health of the system (essentially, looking to classify each increment of the time series). Being biased towards Python, we built datasets and began testing with most of the standard machine learning classification algorithms in sci-kit... Read More »
  • Why python can make your decision making process easier

    Mohsen Moarefdoost Ph.D. Oct 22nd, 2015
    Andy Picture MT 3 Web
    If you are dealing with complex, large scale and multi layered systems, you may end up analyzing and solving multiple, even nested, optimization problems (LP- linear programs, MIP- mixed integer programs, QP- quadratic programs, etc.) in your decision making process. So, you need to design an analytical algorithm that calls optimization problems multiple times, use... Read More »
  • iStock_000043210746_Large Paint web
    At CSCMP’s annual meeting, I got the chance to listen a talk from Benjamin Moore on how they redesigned their supply chain.  Here were the top five lessons I gleaned from the talk (you can find this list and more details in the full article on Supply Chain Digest): Expect a study to take 3-4... Read More »
  • Two Supply Chain Design Lessons from Starbucks CEO

    Michael Watson Ph.D Oct 19th, 2015
    iStock_000017053046_Large Starbucks Web
    The Starbucks CEO, Howard Schultz, spoke at the annual CSCMP event in San Diego a few weeks ago. He made two key observations about the the supply chain: First, you won’t be able to scale your business unless your supply chain can handle it.  (As a side note, he also mentioned that if your HR department... Read More »
  • NFL Fans on Facebook

    Cario Lullo Oct 8th, 2015
    iStock_000041857676_Large Steelers Stadium Web
    In 2013, Facebook conducted an exploration of the number of Likes received by each NFL team in every U.S. county for each day of the season. The study conducted by Sean Taylor involved approximately 35 million U.S. account holders who Liked a page for one of the 32 teams in the league can be found... Read More »
  • CSV to Spark SQL tables

    Diego Klabjan Jul 20th, 2015
    iStock Data Image Sized for Blog
    Recently we were involved in a project that required reading and importing more than 30 csv files into Spark SQL. We started writing scala code to ‘manually’ import file by file, but we soon realized that there is substantial repetition. As a result we created a nice helper object that takes as input information about... Read More »
  • iStock_000041990766_Large Chicago Skyline for Web
    Yesterday’s Chicago Tribune ran an article on how All State and the City of Chicago are combining existing data sets with predictive analytics to solve long-standing problems.  These problems include how to best conduct restaurant inspections to minimize food-borne illness, how to inspect elevators to prevent problems, and which trees to trim to minimize damage after... Read More »
  • iStock_000026778700_Large Multiple Targets for Web
    (This article first appeared in Supply Chain Digest) Let’s start with an example of multi-objective optimization.  The map you see is all the towns and cities in Germany where your product is sold.  Let’s assume you want to to determine how many warehouses you need so that you are within 75km of every customer. This... Read More »
  • iStock_000020359668_Medium Man and Mountain sized for Web
    (Note:  This article first appeared in Supply Chain Digest). A few months back, the Wall Street Journal reported that Ceasar’s “big-data customer loyalty program” was valued at $1B.  This highlighted again the idea that data is becoming an ever more valuable asset. A good book, Big Data, discusses how firms are realizing that that the... Read More »
  • Ford’s $100M Inventory Solution

    Michael Watson Ph.D May 26th, 2015
    iStock_000056977124_Large Autos For Web
    Several years ago, I worked on a technical inventory optimization project for a $1.5B consumer products company.  The CEO of this firm attended a few of the technical meetings where we were talking about things like lead-time variability and the forecast errors.  I asked him why he was so interested in this inventory project. His... Read More »
  • iStock Data Image Sized for Blog
    One of the biggest hurdles in MapReduce is model calibration for machine learning models such as the logistic regression and SVM (Support Vector Machine). These algorithms are based on gradient optimization and require iterative computations of the gradient and in turn updating the weights. MapReduce is ill suited for this since in each iteration the data... Read More »
  • Which Final Four Game Was Most Exciting?

    Bradford Winkelman Apr 9th, 2015
    iStock_000040227242_XXXLarge for Web
    Which Final Four game was the most exciting? The question is difficult because any person who has an opinion is likely heavily influenced by personal loyalties. One objective way of answering the question is to conduct sentiment analysis of Twitter data. The plots below show the total number of tweets per minute* throughout each of... Read More »
  • Elite 8 Game Louisville Michigan State
    Last Sunday we followed March Madness on Twitter to see what fans and rivals were discussing during the final day of the Elite Eight competitions. Looking at the trends you can almost feel the emotions of the various fan bases. For example, check out the spike for #louisville and compare it to the subsequent spike... Read More »
  • NCAA_March_Madness_logo
    We are collecting Twitter data for the NCAA Tournament. We’ll have more to come, but you will like this dashboard. It shows which team has the most tweets (over the last 48 hours- before the round of 64, but during the play-in games)– you might be surprised. You can see whether teams are getting negative... Read More »
  • So what can companies expect to derive from social media, in this case Twitter? To give a little background, Twitter is an online social networking service with more than 100 million users. Users can post and read “tweets” which are short 140 character messages. These tweets are little messages that companies can hone in on... Read More »
  • iStock_000031281272_Double for Web
    What is the status of the truckload transportation market?  Does demand outstrip supply?  Does supply outstrip demand?  Is the market relatively balanced?  These are hard questions to answer because, unlike commodities such as corn or beef, there is no centralized exchange for truckload transportation. So if a shipper wonders why their contracted carriers are rejecting... Read More »
  • opex_contact
    After receiving a lot of feedback on our video interview on data cleaning, we just published an article on Supply Chain Digest on the top five rules for cleaning data in a strategic project: #1:  Be patient. Usually by the time the project starts, the management team wants (or has been promised) fast results. It... Read More »
  • Customer Behavior from Web and Text Data

    Diego Klabjan Mar 3rd, 2015
    Many sites and portals offer text content on web pages. For example, news aggregators such as The Huffington Post or Google News allow users to browse news stories; membership-based portals focusing on a specific industry, e.g., constructionsupport.terex.com for construction, offer members a one-stop page for the latest and greatest updates in a particular domain; in... Read More »
  • a-shot-to-remember-wiki-commons
    Airlines have been one of the first industries to use advanced analytics in areas such as revenue management. “Should a request for a seat be granted?” is a fundamental challenge in the industry. If it is granted, the money can be left on the table since the next day a highly valued business passenger might... Read More »
  • Big Data: A Misnomer

    Diego Klabjan Feb 20th, 2015
    Google Trends shows no organic searches for term “big data” until 2011 and an approximately 7-fold increase in the next two years, and then searches doubled from 2013 to 2015. Google projects further increase in the following years albeit at a lower rate. Google searches for Hadoop, the most popular software for handling big data,... Read More »
  • Do You Have Clean Data for a Strategic Project?

    Michael Watson Ph.D Feb 18th, 2015
    When doing a strategic analysis, most firms either assume they have exactly the right data.  That is, these firms have ERP systems full of data, data warehouses full of data, and seem to have the data they need to run their business. However, the data collection and cleaning process for this type of study always... Read More »
  • iStock Data Image Sized for Blog
    We recently worked on a client engagement that included web data and other customer specific information. It was a propensity analysis type project where recommendations were required for each individual client based on his or her past actions on the web. Each item recommended has many features and clients belong to organizations, which creates interactions... Read More »
  • Since many of the most publicized examples of advanced analytics consider companies in the private sector, we don’t often think about how these techniques apply in humanitarian and nonprofit organizations. Yet, the public sector provides exciting applications of analytics that have a positive impact on the community. For instance, the American Red Cross, Greater Chicago... Read More »
  • Recently Hortonworks published the attached white paper on how the advent of big data technologies have given rise to a “new ultra-competitive breed of business that consumes the full spectrum of its data transforming immense volumes and varieties of data into currency.” This paper addresses how these “Data First” enterprises are investing in advanced analytics to garner... Read More »
  • Northwestern Professor and Partner at Opex Analytics, Diego Klabjan recently authored the following column on same-day delivery and its impact on the design supply chain in Supply Chain Digest. The article highlights past failed attempts of a decade ago of firms like Webvan and Kozmo with recent efforts by companies ranging from start-ups such as Deliv... Read More »