MindSet Predictor is a versatile solution for numerical/discrete prediction, for small to large problems.
MindSet Predictor scales with your data, by working from files, SQL database or connecting to your application.
MindSet Predictor can take into account dozens of variables, by using state-of-the-art variable selection techniques.
As more data becomes available, MindSet Predictor adds more explanatory variables and adjusts for the resolution at which data is analyzed.
Define models for time series, with continuous or discrete variables.
Define user rules for filtering data.
Allows several models to run simultaneously from a single computer, with time schedules for prediction.
No knowledge of mathematical modelling required.
MindSet Predictor can be used as desktop application with a supplied GUI, or deployed with .Net Application.
MindSet Predictor performs automatic variable selection over the data set. As more data becomes available, more variables are inserted, by relevance. The generated models are robust to noisy and highly non-normal data.
MindSet Predictor is a complete suite, building on known prediction and variable selection algorithms such as: Regression Trees, Lasso, Robust regression and classic variable selection methods.
MindSet Predictor is able to compensate for noisy data and non-normal noise distributions by utilizing non-parametric tools and custom data filters. These methods make MindSet Predictor a versatile solution for many types of datasets.