|
 |
| Predicting Insurance Cross Sell
|
We present empirical evidence of this approach as compared to other machine learning methodologies when applied to real-life insurance policy data sets. We also demonstrate how our non-linear based interpretability help to explain the key indicators affecting the prediction as well as discover hidden business trends and rules that intuitively explain the behavior.
|
|
 |
| Intensive Care Study
|
In this case study, we present empirical evidence of this approach when applied to patient mortality after a stay in the Intensive Care Unit (ICU) . We also demonstrate how non-linear based interpretability help to explain the key indicators affecting the prediction as well as discover hidden business trends and rules that intuitively explain the behavior.
|
|
 |
| Predicting Compound Activity against HIV
|
Equbits Foresight has been applied to the NCI HIV data set. The classification results were compared to existing classification algorithms to demonstrate superior performance, accuracy, and were validated to demonstrate generalization without overfitting.
|
|
|
| SAS Enterprise Miner-Equbits Foresight Case Study
|
This case study shows how Equbits Foresight yielded significantly superior results when compared with SAS Enterprise Miner (Logistic Regression, Neural Nets, Trees). This case study was jointly developed by SAS and Equbits Inc. for 2005 SAS User Group Conference (SUGI).
|
|
|
|
| Predicting Central Nervous System/ Blood Brain Barrier Permeability
|
Equbits Foresight has been applied to a set of known compounds that are CNS positive and negative. The classification results were compared to existing classification algorithms to demonstate superior performance and accuracy.
|
|
|
| Case Study- Calculation of Aqueous Solubility of Compounds
|
Equbits Foresight has been applied to a widely known solubility dataset. The regression results were compared to existing regression techniques to demonstrate superior performance, accuracy, and were validated to demonstrate generalization without overfitting.
|
|
|
| Local Modeling of Mutagenic Compounds
|
This case study demonstrates the effectiveness on local modeling of a known chemical data space for mutagenic compounds. Equbits Foresight demonstrates improved accuracy compared to traditional modeling methods. Advanced model and predictive results interpretation techniques are presented and discussed.
|
|
|
|
|
|
Case Studies
News & Events
