Features
Neuton is Currently
Featuring
Automated creation of the neural network structure to utilize highest predictive power possible
Automated training and growing of the model with the best characteristics
Automated dataset preparation (preprocessing), including text features and time series
Automated feature engineering
Predictions via Web interface
Predictions via REST API (including Scala, C Sharp, Java and Python)
Downloadable models for local use without Neuton
Embedded model capabilities
GPU Support for training
Neuton Explainability Office ®
Exploratory Data Analysis
Feature Importance Matrix with class granularity
Model Interpreter
Feature Influence Indicator
Validate Model on New Data
Model-to-Data Relevance Indicators historical and for every prediction
Model Quality Index
Confidence Interval
Expansive list of supported metrics
Model Quality Diagram
Problems Neuton
is Currently Solving
Regression. Prediction of continuous numeric value (Ex.: temperature or real estate prices)
Time series. Prediction of numeric values dependent on the date and time (Ex.: sales prediction)
Binomial classification. Prediction of one of the two outcomes (Ex.: ‘Yes’/’No’, ‘White’/’Black’, 1/0)
Multinomial classification. Prediction of a value that may have 3 or more categorical outcomes (Ex.: next best offer prediction from 10 categories of products)
Positioned
for the Future
Support of Image/Video
Multi host training
Stay updated, join the community
slack