8.1 CRISP-DM: Data Mining Process

This initial phase focuses on understanding the project objectives and requirements from a business perspective, and then converting this knowledge into a data mining problem definition, and a preliminary plan designed to achieve the objectives. A decision model, especially one built using the Decision Model and Notation standard can be used.

The data understanding phase starts with an initial data collection and proceeds with activities in order to get familiar with the data, to identify data quality problems, to discover first insights into the data, or to detect interesting subsets to form hypotheses for hidden information.

This phase is where most of the course has been focused on so far. Visualizing data in Tableau, creating clusters, testing for non-linearities, normality, and heteroskedasticity are all techniques used to understand the data. The vast majority of your time will, and should, be spent on understanding the data. Because if you don't understand the data, then you are taking large risks when it comes time to perform modeling and make business decisions on those models.

The data preparation phase covers all activities to construct the final dataset (data that will be fed into the modeling tool(s)) from the initial raw data. Data preparation tasks are likely to be performed multiple times, and not in any prescribed order. Tasks include table, record, and attribute selection as well as transformation and cleaning of data for modeling tools

Recall tasks we've performed on the Bike Buyers and Lending Tree data to change variables from categorical to ordinal, make logarithmic transformations, and to calculate exponential variants are all examples of data preparation. Typically, the data scientists will initially perform these transformations manually until they are sure of their results. Then, the cleaning and preparation will be automated into an ETL process.

In this phase, various modeling techniques are selected and applied, and their parameters are calibrated to optimal values. Typically, there are several techniques for the same data mining problem type. Some techniques have specific requirements on the form of data. Therefore, stepping back to the data preparation phase is often needed.

This phase is where much of the remainder of this course will focus. We will learn many different forms of modeling and alternative calculations for each modeling type. In the previous chapter, the "modeling" phase was the part where you analyzed regression models which produced coefficients (a.k.a. "weights") that explained exactly how important each varible was.

At this stage in the project you have built a model (or models) that appears to have high quality, from a data analysis perspective. Before proceeding to final deployment of the model, it is important to more thoroughly evaluate the model, and review the steps executed to construct the model, to be certain it properly achieves the business objectives. A key objective is to determine if there is some important business issue that has not been sufficiently considered. At the end of this phase, a decision on the use of the data mining results should be reached.

How was this step performed in the last chapter? Well, we started some prelimiary evaluation when we examied R squared, which indicated how well the regression model variables (age, income, commute distance, etc) expalined the dependent variable (purchased bike)

Creation of the model is generally not the end of the project. Even if the purpose of the model is to increase knowledge of the data, the knowledge gained will need to be organized and presented in a way that is useful to the customer. Depending on the requirements, the deployment phase can be as simple as generating a report or as complex as implementing a repeatable data scoring (e.g. segment allocation) or data mining process. In many cases it will be the customer, not the data analyst, who will carry out the deployment steps. Even if the analyst deploys the model it is important for the customer to understand up front the actions which will need to be carried out in order to actually make use of the created models.

In the bike shop example, we deployed the model by building the Excel-based calculator that made use of the model results. However, that is a very limited deployment because we would then have to make that Excel file available to others and have a mechanism to update the model occasionally as more data is collected. In practice, we integrate model results into information systems (e.g. think Amazon's prediction of which products you'd like based on what you've already clicked on). The information system not only uses the model results to guide customer/user behavior, but also records the new behaviors (e.g. purchase decisions) which are fed back into future iterations of the model: thus, completing the loop required for machine learning to occur.