The Ultimate Modeling Software
A complete suite of data analysis, modeling and scoring tools – in one easy to use environment.
Effective targeting of prospects and customers is critical for the success of a direct marketing program. In Model Miner, you can quickly and effectively target and promote to your best prospects and customers with an easy-to-use and powerful modeling software.
Model Miner gives you the ability to control directly the entire process of profiling, model building, scoring, and targeting of your best prospect and customer segments. You will be able to keep your valuable data in-house and protect the confidentiality of information that is responsible for the success of your marketing program.
Model Miner gives you the analytical capabilities that set you free from dependence on high-priced outside experts and reduce your in-house research costs.
Model Miner’s Features:
- Built By Professionals and Experts
Model Miner was designed, built, and tested by a team of experienced marketers, sophisticated statisticians, and seasoned software engineers.
Model Miner provides a self-training wizard that, along with a help manual, enables you to quickly learn and continuously sharpen your ability to construct powerful models.
- Expert System with Cross-training procedure
Model Miner uses built-in decision rules to drive the exploratory data analysis (EDA) and the model building procedures. Model Miner automates all data pre-processing, variable construction and testing, and model estimation functions.
- Model Miner automates data hygiene functions that eliminate bad data, remove outliers, and correct for missing values. It performs mathematical data transformations and date conversion. It also automates the front-end exploratory data analysis procedures, and automatically groups and tests all possible combinations of data categories to derive significant new variables.
- Model Miner’s boosting methodology cross trains multiple techniques in model construction, including CHAID, linear regression, logistic regression, weighted regression, probability neural network, and back propagation neural network. For building response, cross sell, retention, and reactivation models, for instance, Model Miner uses a combination of CHAID analysis, stepwise logistic regression analysis, and probability neural network analysis. Similarly, for building revenue, profit, and AREA models, Model Miner cross trains a combination of linear regression, weighted regression, and back propagation neural network analysis.
Model Miner displays tables and graphs for every step of the modeling process. Reports are provided for frequency analysis, CHAID analysis, bivariate analysis, stepwise regression analysis, neural network analysis, and scoring. Model Miner also creates a number of diagnostic reports for assessing model performance.
Model Miner constructs a model in seconds or minutes. For small data sets, Model Miner will construct a model, create reports, and score file within seconds. For larger data files, Model Miner will construct a model, create reports and score file within minutes.
Model Miner can score a file internally or export a SAS program for scoring a file externally in your database.
Model Miner’s Benefits
Model Miner is quick to master and easy to use. With Model Miner’s self training wizard, you will be able to build powerful models following a one-day training session.
Model Miner will significantly reduce model construction time. By automating the variable construction and testing procedures, Model Miner will eliminate 95% of an analyst’s front-end number crunching and testing effort. This will result in significant savings in model building time and effort.
- Guaranteed Optimum Models
Because Model Miner creates and tests all possible combinations of variables before it constructs a model, it will leave no possibility un-examined. As a result, Model Miner will always identify and capture the best combination of variables its models. The models will be optimum and most powerful for any data sample analyzed.
- Clear-cut Back-end Decision Process
Model Miner’s cross-training methodology streamlines back-end decision efforts and offers clear cut solutions. Analysts are no longer required to choose among alternative models. Model Miner will automatically identify the best model.
Model Miner creates easy-to-interpret Crystal Reports for every step of the modeling procedure. The reports are exportable to a number of reporting platforms, such as Word, Excel, Adobe, etc. This gives you the ability to customize each report.
With Model Miner, you can perform extensive “what-if” analysis. You can easily test alternative models and take effective steps to ensure a model’s consistency and reliability.
Model Miner’s Applications
Model Miner constructs a range of models predicting behavioral propensity. These models identify the best prospects and single out the best customers according to their response, cross sell, up sell, retention, reactivation, and risk propensity. With the aid of a model, you can create and implement optimum contact strategies for your marketing campaigns.
Model Miner constructs revenue and profit models. These models identify the potentially profitable prospects and isolate the top performing customer segments that are responsible for producing 80% of revenue and profit.
Model Miner constructs area models. These models predict potential sales and response rates in geographic areas, such as ZIP Codes, for site selection and list rental applications.