Walmart processes over 20 million point-of-sale transactions every day. This information is stored in a centralized database, but would be useless without some type of data mining software to analyze it.

If Walmart analyzed their point-of-sale data with data mining techniques they would be able to determine predictable sales trends, develop customized marketing campaigns, and more accurately predict customer loyalty and improve retention.

CVS Health and the call center. CVS launched a call center pilot program by analyzing speech patterns, Corp. matches call center employees with customers based on personality type to improve interactions and ultimately boost sales and customer retention.

CVS began experimenting with "predictive behavior routing" by segmenting customers into about six behavior groups. The drugstore chain is also scoring call center agents so that it can match agents with the kinds of customers they'll best engage with. The pilot reduced call time and improved the performance of the calls themselves, and was successful.

Fast food and video. This company is training cameras on drive-through lanes to determine what to display on its digital menu board. When the lines are longer, the menu features products that can be served up quickly; when the lines are shorter, the menu features higher-margin items that take longer to prepare.

Wal-Mart Stores Inc. and search. The mega-retailer's latest search engine for Walmart.com includes semantic data. Polaris, that was designed in-house, relies on text analysis, machine learning and even synonym mining to produce relevant search results. Wal-Mart says adding semantic search has improved online shoppers completing a purchase by 10% to 15%. In Wal-Mart terms, that is billions of dollars.

Connection Analytics helps us find patterns in the way people and things of all types are related. These connections are being used for business value.

Suppose that we are interested in the factors that influence whether a political candidate wins an election. The outcome (response) variable is binary (0/1); win or lose. The predictor variables of interest are the amount of money spent on the campaign, the amount of time spent campaigning negatively and whether or not the candidate is an incumbent.

Both analytics and data mining are aimed at information that is actionable; however, analytics and data mining are as different as chalk and cheese. Data analytics is distinguished from data mining by the scope, purpose and focus of the analysis.

Analytics usually comes with hypotheses testing. The analyst has something in mind and is looking to answer a question and has a hypotheses about that question.

Data miners sort through huge data sets using sophisticated software to identify undiscovered patterns and establish hidden relationships.

Classification is supervised learning technique of grouping the data, that uses a predefined training data to define a class and then group on basis of that. Classification model is created from training data, then the classification model is used to classify new instances.

Clustering is unsupervised technique of grouping the data, without using any predefined training data . The clustering algorithm is supposed to learn the grouping on the fly.

In classification, you first 'Learn' what goes with what and then you 'Apply' that knowledge to new examples. So if somebody gave us the first picture on the left, which is a plot of hair length (Y axis) against gender (on X axis, however sorted such that the points belonging to females, corresponding to blue color, appear first, followed by male, corresponding to red color), the task of a classification model would be to learn the fact that typically females have longer hair than males and then use this knowledge and apply it to graph (obtained from different sets of people) shown on the right where there is no color coding done. A classifier then has to look at each black point, see its Y axis value and from the knowledge it acquired from left graph, guess if it should be blue or red.

A clustering, on the other hand, is that kind of classification where you never get a chance to see color coding, i.e. something like the plot shown below:

In this case, clustering algorithm has to "Infer" that you could create at least two groups of points. Now its beyond clustering algorithm to put names to each group i.e. after creating the two groups, clustering algorithms cannot tell you whether the first group corresponds to males or females.

Data comes in all shapes and sizes: audio responses to an interview or focus group, written responses to an online or paper questionnaires, diary entries, documents, observation notes etc. Much of this data is in Words/Text and Numbers , or can be turned into Words/Text or Numbers .

Generally, data is split into two types:

Quantitative:Numbers, and other data that can be transformed into numbers

Qualitative:Words/text and other data which come in a non-numerical form.