|
Neuton (network)
|
|
Neuton is a neural network framework and automated machine learning (AutoML) solution developed by Bell Integrator. It is based on its own proprietary algorithm protected by a patent issued in 2019. Overview Neuton allows users to build, train and deploy neural network models to solve problems of multivariate regression and classification. Neuton is available as an enterprise solution to run on-premises or an online service on Google Cloud Platform. Predictions are made via a REST API or Web interface. The resulting model can be used as a service deployed in the cloud or downloaded for local use. Benchmark tests verify that Neuton outperforms competitors by several orders of magnitude. Features Neural Network Framework *Automated Neuton model architecture creation *Automated hyper parameters configuration *Regression and Classification problem solving *No limitations on dataset size for training *No overfitting *Automated training completion upon reaching optimal model and predictive power while avoiding overfitting Auto ML *Automatic and manual problem type definition (regression or classification) based on dataset analysis *Automated dataset preparation (preprocessing) *Automated feature engineering *Automated training *Automatic training completion upon user input thresholds reach *CSV datasets support Preloaded datasets *Models quality metrics visualization. Predictions *Capability to download trained models in HDF5 format with Python API for local usage along with script examples on how to use Neuton models offline *Web Interface for predictions on trained models *Predictions via REST API *No limitations on input data size for predictions Infrastructure *Cloud Based solution *SaaS solution *GPU Support for training *Automatic and seamless provisioning and release of infrastructure necessary to perform training and web hosting for predictions 
|
| < Prev | Next > |
|---|
Main Menu
| Home |
| Articles |
| Search |
| Submit Article |
| Contact Us |
| Donate to Wikibin |
