Login Register

Auto-login on future visits

Forgot your password?

Paint labels by hand.

Labeling samples by hand saves your time. Instead of fiddling with cluster analysis to identify that specific class mode, simply paint it. Removing outliers has never been easier.

Interactive painting of sample labels in perClass.
Inspecting meta-data of samples such as multiple labels

View meta-data of your measurements.

Your problem is more than data matrix and sample labels. There are often very useful meta-data such as patient ids, object labels or frame numbers. Easily include meta-data in you data sets. Use them to get subsets, group decisions or perform leave-one-group-out evaluation.
See a short video

Live class distributions.

When working with large data sets, scatter plots become too crowded. In such situation, you greatly benefit from live distribution plots directly in the scatter view. Understand the nature of class overlap. Quickly learn about important trends and feature combinations.

Live class distributions in scatter plot
Drawing polygon classifier directly in feature space.

Sketch your classifiers by hand.

Sometimes, drawing a classifier is faster than tediously searching for a good model. Want to discard outliers or separate that group of classes from the rest? Just draw a polygon around it. If you wish, you may still tune it with ROC analysis like any other statistical classifier. So you don't loose your precious target objects.

See a short video