“Opal is Australia’s national gemstone, but no new significant opal discoveries have been made since the early 1900s. Most opal exploration is carved out by individual miners digging in desolate areas around old opal fields in the Great Artesian Basin.”

The University of Sydney’s data mining strategy has turned this around. It has already successfully identified lucrative new opportunities, one of which has already been confirmed as a successful mining location.

This information and technique could prove invaluable to companies like Gemfields who are interested in bringing more order to the gem mining industry. Gemfields has expressed a desire to create a more predictable supply of precious gemstones, especially colored varieties like emeralds.

After all, Gemfields would simply need to apply the same technique to the geologic data that is the most conducive to locating emeralds instead of opals.

In fact, this data mining process seems like it would be absolutely instrumental to any attempts to modernize mining, since it reduces the chances that labor costs and other resources would be wasted on places that aren’t going to be productive.

Having the proper training is crucial to making the most of data mining advances, in this or any other industry. That’s why The Modeling Agency offers a free data analysis webinar and other data analysis training courses. Make the most of these exciting new developments!

The post Data Analysis Helps Gem Miners appeared first on Data Mining and Predictive Modeling: Training & Consulting.

The title references sweeping changes that J.C. Penny’s former CEO Ron Johnson made in a trial-and-error attempt to increase sales. The attempt flopped.

Yet in the past, trial-and-error were all that retailers really had to guide them through the marketplace.

Admittedly, once-upon-a-time that marketplace moved a lot slower. Retailers weren’t competing with e-commerce. Nor were they dealing with customers who could compare prices by the simple expedient of whipping out their cell phones.

In the Information Age, retailers can really only seize a single edge: customer experience.

“For retailers, competing based on price, location, and even product is no longer sufficient. The way they’ll win is by providing a superior customer experience–which is proven to be an accurate predictor of financial performance.”

So what can data do?

Data can help retailers figure out exactly how to provide that experience.

Good retail data analysis can draw from several sources, including customer surveys, point-of-sale data, social media, and all sales channels both offline and online.

Together, these sources work together to identify what customers want. Thus, retailers can create customer experiences that keep customers coming back for more by predicting the outcome of certain changes.

Furthermore, this data can help identify places where retailers are already turning customers off so that they can correct problems. They can also fill gaps in their inventory so they aren’t leaving money on the table.

Of course, all of these benefits depend on having the proper training to harness and reap these benefits. That’s why The Modeling Agency offers a free data mining webinar and other analytics training courses. If you’re in retail you should consider taking advantage of these powerful tools today so that you can seize the edge you need to stay alive in today’s hyper-connected, hyper-competitive marketplace.

The post Data Mining Frees Retailers from Trial-and Error They Can No Longer Afford appeared first on Data Mining and Predictive Modeling: Training & Consulting.

For example, fleet managers can use these systems to determine residual values, which helps them make better decisions about purchasing, leasing, or selling their inventory. Residual value calculations are impacted by a number of different factors, and mistakes can cost companies dearly. The ability to get it right has an incredibly positive impact on managing both costs and cash flow. In fact, this value determination is the first place where the savings really start to add up.

Predictive modeling allows managers to adjust for the external factors that impact their business, factors like interest rates and gas prices that are beyond their control. Getting a clear picture of the ways that these issues will impact the fleet’s upcoming needs and challenges allows fleet managers to shield themselves from risk while keeping the fleet functional.

In order to reap these benefits, however, fleet managers need adequate data analysis training. According to Telogis, the leading producer of fleet management software platforms, “analysis paralysis” is one of the top 8 challenges faced by today’s fleet manager. Telogis estimates that fleet managers receive 36 million packets of data every day (assuming a fleet of 50,000 vehicles equipped with GPS devices). This highlights the need for fleet managers to understand how to deal with that data in a meaningful way, one that lets them make the decisions that matter most to them.

The Modeling Agency can help fleet managers make the most of their data. TMA offers comprehensive data analytics training courses which can help companies make the most of their assets while reaping million-dollar savings rewards for their fleets. Start by making the most of this free webinar, designed to help companies get a better understanding of data mining and predictive modeling.

The post Data Mining Helps Fleet Managers Make Better Decisions appeared first on Data Mining and Predictive Modeling: Training & Consulting.

Although collection of data through electronic means is absolutely viable and, with most big data, absolutely essential, the actual mining of data requires clear direction and individual understanding. For instance, as we know, the very first step in mining data is to determine, not only the challenge or problem to be solved, but most crucially, the question to be answered in order to solve it. This, of course, requires an individual to have a clear business understanding. Oftentimes, determining the question may require a data miner to consult business experts in order to define the clear question.

As described by SmartDataCollective.com, what comes next is “understanding the data, the way they have been collected, their particularities… [and deriving] knowledge useful for preparing the data.” Clearly, as this is a task centered on internalizing data and its absolute meaning, a machine is not up to the task, no matter how advanced or orderly.

Data preparation for mining and analytics can, to a degree, be automated. Even in this case, when raw data is combed through to leave only relevant, useful information, outliers and missing fields could easily stump a computer. Detection of issues may be straight-forward enough, but deciding how to assess and move forward is a task for none but a data scientist. Again, the core to all of the decision making within data mining as a whole is genuine business understanding of the problem at hand. How can we expect a computer to manage this task?

Assuming that “that there are enough data, that the choice of the algorithm is not business dependent (which is usually not the case) and that the evaluation criterion is known,” SmartDataCollective.com tells us, “electing a data mining algorithm and tuning its parameters [through careful testing]  is certainly the task that can be the most easily automated.” What follows, however, is the process of evaluating the results and then implementing a solution to the root question, once again requiring business knowledge and deductive reasoning.

It is clear that data mining in its simplest forms can be automated, but for a business owner who intends to make decisions affecting the future of his or her company, a deeper understanding of what the data means is essential. For an amateur data scientist, this can seem a daunting task. Fortunately, The Modeling Agency offers a free webinar to help you grapple the problem. If you’re ready to dive into this exciting challenge further, perhaps it’s time that you learn about TMA’s intensive Training Seminars.

The post Can Data Mining Be Automated? appeared first on Data Mining and Predictive Modeling: Training & Consulting.

A clear example of this challenge will unfold as we watch the development of businesses like Groupon, whose client base is at the root of its successes. Groupon offers daily online discounts which are highly focused to a client’s georgraphic placement and their interests. At once a novel concept and one which is no longer unique, Groupon’s business model, though sound, will be challenged by imitators, especially larger competitors with vast stores of consumer data to mine, like Amazon. Amazon, a super power in online commerce, has been tracking and storing data on consumer purchase history and even browsing habits for some time. How can a smaller company like Groupon hope to compete? As described in an editorial on the Financial Times, James Bateman writes,

“My feeling is that Groupon is worth the value of its email database, and if it has data on customer spending habits (that is, which promotions they have responded to) then for many businesses this could theoretically be a value acquisition. But trying to compete against Amazon – if the latter takes the daily-deal market seriously – seems doomed to failure. The only consolation that I can offer Groupon at the moment is that Amazon’s focus seems poor.”

Mr. Bateman cites Amazon’s failure to specify more carefully which clients receive which deals as Groupon’s key to survival. Given enough time to grow and cultivate their consumer list, could Groupon compete realistically with a giant corporation like Amazon? Perhaps not, but we can clearly see that it is data mining and predictive modeling of client behaviors that is allowing Groupon to hold on against this giant adversary.

Whether your company is competing in the vast, online marketplace as well, or has a more direct relationship to your clients, data mining can be a key factor in your growth and success as well. Saving you from wasting resources and energy on misplaced or misguided marketing efforts, Data mining can help your business identify new outlets and opportunities.

If you are ready to learn more about how data mining can support your marketing efforts, sign up for one of The Modeling Agency’s free webinars. To gain a broader understanding, TMA also suggests attending their in-depth Training Seminars.

The post Data Mining is Key to Survival in a Global Marketplace appeared first on Data Mining and Predictive Modeling: Training & Consulting.

However, the payoff is multitude for the business owner who invests the time and energy into collecting and mining accurate data from the beginning. Data mining all of the information collected from previous transactions will give a startup a clear view of what exactly has happened so far– and how, then, to repeat, or improve, the process. This is where predictive analytics comes to play. For a number of tasks, modeling future potential outcomes, though perhaps time-consuming at first, will save a small business owner any number of headaches. The resources which may be preserved through this process are plentiful. So much so that Forbes calls this process a ‘goldmine’, and shares a creative list of examples, including some of these highlights:

Targeted Direct Marketing

Why waste money on Marketing or Advertising campaigns that won’t reach your target? Rather than relying on trial and error, Data mining allows a small business owner to extrapolate exactly who his consumer is and how he can best focus his efforts on reaching them with the least investment and highest return.

Customer Retention

Once your company initiates a client relationship, nurturing a positive affiliation is crucial. Determining which clients may be about to leave and for what reason would save a business owner’s energy for only focused efforts to correct errors and maintain a cohesive working relationship.

Investment Risk Management

From purchasing stocks to vetting partners or acquisition candidates, even to selecting which vendors to trust with your business, predictive modeling can save a business owner from potentially costly business errors. Analyzing the data that has come before can help a small business side-step the pitfalls.

Education, Fraud Detection, Political Campaigning, Mortgage Underwriting, and many more outlets… The application of predictive analytics will vary in each field, but the same truth is evident throughout. The investment in gathering and mining your data to allow you the ability for predictive modeling is of tantamount importance. If you are ready to learn more, consider taking advantage of The Modeling Agency’s Free Introductory Webinars. For a fuller data mining experience which will leave you with enough practical knowledge to embark on data mining for your small business, TMA’s Training Seminars may be the right answer for you.

The post Data Mining for Startups appeared first on Data Mining and Predictive Modeling: Training & Consulting.

Firstly, a Horizontal Data Scientist is one who has a broad knowledge. Whether computer scientists, statisticians, software engineers, or even predictive modelers, these data scientists can extract copious information from their given field. This type of data scientist values speed and is able to extract data while anticipating errors and weeding out false positives. Their data sets may be flawed, but they have a high success rate of pulling usable information from these sets because of their real world thinking. Horizontal Data Scientists are highly skilled at establishing exactly the question that should be answered, as well as shaping the correct structures and metrics to get there. Often, this type of knowledge comes from experience rather than in-depth knowledge of theoretical conceptions.

On the other hand, we have Vertical Data Scientists, those who “combine vision with technical knowledge,” to quote Mr. Vincent Granville of DataScienceCentral.com, “They might not be expert in eigenvalues, generalized linear models and other semi-obsolete statistical techniques, but they know about more modern, data-driven techniques applicable to unstructured, streaming, and big data.” Vertical Data Scientists are the type who are sought out by recruiters, those who come from university programs. Very often, however, university programs in ‘data science’ tend to strive to train either a computer scientist, a statistician, an operations research or a MBA guy, but seldom any combination thereof. Therefore, the big complaint with this type of Data Scientist is that he is often too ‘by the book’, trusting theories over pragmatism and pulling from real-world practicality.

The Modeling Agency is proud that their entire bench is made up of a rare and highly seasoned team of horizontal senior consultants, those who experience is proven by results. Which type of data scientist do you intend to be? If you’d like to learn more about how The Modeling Agency’s training programs can help you see the full picture, read about TMA’s Training Seminars. Not sure you’re ready to dine in? TMA also offers free introductory Webinars.

The post Differentiating Data Scientists appeared first on Data Mining and Predictive Modeling: Training & Consulting.

The challenge was certainly worthwhile in the case of a group of scientists from Microsoft, Stanford, and Columbia University. For the first time, unreported side effects of various medications were able to be detected and classified before the Federal Drug Administration’s warning system could find them.

By using automated software, the scientists were able to troll millions of internet queries for similarities. The study focused in particularly on an antidepressant and a cholesterol-lowering drug, which were ultimately found to cause a dangerous spike in patient blood sugar upon interaction. Prior to this study, the only way such an interaction would be noticed would be if a physician detected the result in a patient and submitted their finding through the FDA’s Adverse Event Reporting System.

This project came to be when Russ B. Altman, chairman of the bioengineering department at Stanford, began work to automate searches of drug-drug interactions within the FDA’s reports. Upon identifying the interaction within that source, Altman wondered if there might not be a more accurate, faster way to make such connections. And so, he turned to Microsoft.

Microsoft engineered a software toolbar to scan anonymized data collected from web users who allowed their browser information to be shared. Through this method, the group was able to analyze 82 million individual searches for drug, symptom, and condition queries. After continually refining the searches, the group was able to glean that those individuals who searched for both of the drugs in question within a twelve month period were much more likely to search for symptoms indicating a hypoglycemic reaction.

As reported by the New York Times, the researchers were surprised by the strength of the “signal” that they detected in the searches and suggested that methods like this one would be a valuable tool for the F.D.A. to add to its current system for tracking adverse effects. “You can imagine how this kind of combination would be very, very hard to study given all the different drug pairs or combinations that are out there,” said Eric Horvitz, a managing co-director of Microsoft Research’s laboratory in Redmond, Washington. “There is a potential public health benefit in listening to such signals,” they wrote in the paper, “and integrating them with other sources of information.”

The FDA has begun to undertake similar methods of research, hoping to study the behaviors of up to 100 million American pharmaceuticals customers. Thanks to innovative and careful data mining, the FDA now has the tools to garner new information which will help to protect the American public from currently unknown risks.

If you’re interested in learning more about how Data Mining can help you unlock hidden information in your own field, consider joining The Modeling Agency for a free webinar. Or, for a more in-depth experience that will leave you with fuller mastery of this valuable tool, perhaps one of TMA’s training seminars is for you!

The post Data Mining Uncovers Drug Interactions appeared first on Data Mining and Predictive Modeling: Training & Consulting.

This tale is a wonderful example of how data means nothing unless one is able to deduce meaningful connections and draw helpful information from it. The protagonist in our film example, for instance, recognizes that players with less experience are predictably less expensive, but, based on reliable statistical data, can be expected to play just as well as more experienced players. Further, one must determine which pieces of data are informative and which are misleading; in our example, RBI (runs batted in) totals, prized by most general managers, were deemed deceptive since, in order to have high RBI stats, other players must have first arrived on base ahead of the given player. Therefore, the most influential statistics in showing a player’s ability to score runs were decidedly on-base and slugging percentages. Such careful extraction of what the data is truly telling the individual is the heart of effective data mining.

As the story goes, in 2002, this method of careful deduction helped the Oakland Athletics to finish first in the American League West with an unexpected record of 103 wins and 59 losses for such a low-budget team. This exceptionally chosen group also set an American League record of winning 20 consecutive games.

As with any case of data mining, however, how the data is considered and compared with other elements determines its true value. “Statistics themselves can be very misleading,” writes Osama K. Solieman, in a book called Sports Data Mining, “Certain players are able to build impressive stats but have little effect on a game. On the other hand, there are players who make a significant impact on the game without having impressive statistics.” These statistics and the perception thereof has a strong effect on how players are rewarded for their efforts on payday, giving them great incentive to make plays which may affect their teams less and their personal statistics more. In this truly rich read from Spider Magazine, writer Salman Haqqi outlines how data mining influences a variety sports, including sports media attention and the multi-million dollar business thereof.

As Haqqi sums this lesson up, “Data and analytics [is] helping athletes and sports organisations improve their performance and take sports to new levels… The scary part of this field of analytics is that it is still in its infancy… Its usage in sports, with its neatly defined parameters, has been a perfect place to realise its practical potential [for use within other fields]. And time will reveal the endless possibilities it holds for the future.”

If you’re interested in learning more about how Data Mining can help you unlock hidden information in your own field, consider joining The Modeling Agency for a free webinar. If you’re ready to master this valuable tool more fully, perhaps one of TMA’s training seminars is for you!

The post A Scientific View of the Game appeared first on Data Mining and Predictive Modeling: Training & Consulting.

Firstly, as analytics become more advanced, some important terms are emerging to reference new, narrowly focused, and highly specialized tools and technologies. As identified by VentureBeat.com, the top items about which you’ll want to be in the know are as follows:

1. Smart Data: As companies needs to more efficiently mine data through solely electronic needs takes precedent in the coming years, one can expect to hear more and more about ‘Smart Data’, data which can be processed without a human mind combing through it by utilizing predictive analytics to anticipate consumer actions. In today’s marketplace, examples such as automated personalizations and recommendations through companies like Amazon, Netflix, and LinkedIn can already be clearly observed.

2. Data Science: A useful term whose popularity has skyrocketed recently, with close to but perhaps not so much overuse as is associated with ‘Big Data’, ‘Data Science’ is meant to refer to a new field which utilizes statistics, machine learning, natural language processing, and computer science to extract meaning from large amounts of data, often with the goal of creating new data products.

3. NewSQL: The scalability of NoSQL combined with the strong ACID guarantees of legacy relational databases offers users new options when dealing with relational data. NoSQL will continue to be valuable for companies who do not require an ACID guarantee, but NewSQL is a solid buzz word of which to be aware and an item whose presence will likely continue to grow.

4. Predictive Analytics: A complicated process which relies on advanced machine learning and statistics to recognize and exploit patterns, predictive analytics is perhaps the root of many of the above mentioned items. Relying on manipulation of historical data to anticipate future actions allows Data Scientists to advise companies and consumers in their best interest. Activities in this arena are applicable in any possible industry and are the driving force behind consumer recommendation services and even fraud detection.

Although the circumstances that gave rise to ‘Big Data’ are still relevant, processes in this arena have progressed, and will continue to progress, beyond the need to store copious amounts of data. New and better systems for processing data will continue to emerge, as will more highly specific terms to describe them.

If you’re interested in learning more about Data Mining’s changing landscape, consider joining The Modeling Agency for a free webinar.  TMA’s training seminars are another, more in-depth way to canvas this growing field and ensure that your company is receiving the maximum benefit from your amassed data.

The post ‘Big Data’ Is Dead appeared first on Data Mining and Predictive Modeling: Training & Consulting.