Retalon, an award-winning AI & Predictive Analytics software provider announced a new approach to training of Neural Nets, named Progressive Learning.
Traditionally data used to train neural net models is less than perfect, resulting in errors and biases. This is not a new problem, as bad data has always been an issue, independently of what technology was used. With Neural Nets, however the idea was that the Neural Nets will at least partially replace data scientists in data preparation process. The same way a good Data Scientist can look at the data and say: “it doesn’t seem to be right”, a properly trained Neural Net was hoped to do the same.
Retalon has developed a new process that allows NN to start training on the trustful data first, gradually learning from less perfect data. The new approach also affects the initialization step. Retalon Progressive Learning technology offers one consistent gradual approach to initialize and train NN for business-specific application.
“You don’t show a 2-year-old the “Game of Thrones” to educate them on how this world works. You start with “Sesame Street”, and then add complexity to the established foundation. We found that this is also an important step in Deep Learning of artificial systems. At the end of the day, all systems (whether human or artificial Neural Net) are based on the same principles. Progressive Learning technology bridges this gap in the process of initialization and training of Neural Nets for business specific applications”, said Mark Krupnik, Ph.D., CEO at Retalon Inc.
Read More : SalesTechStar Interview with Scott Lasica, Chief Sales Officer at Stream
The Retalon Progressive Learning approach offers companies the advantage of training models that will be more tailored to their business process without overfitting. The approach has already shown significant improvement in quality and stability of results in situations with missing or wrong labels, incomplete data, and presence of outliers. Retalon’s new Progressive Learning technology automatically identifies at least 80% of anomalies in data, and requires much less data scientist intervention, time, and resources.
Read More : Three Ways to Measure Whether Sales and Marketing Teams are in Sync