Deep Learning Recommendations

Portable – efficient- scalable

 

maxML is a Deep Learning solution stack for Recommender applications on Time Series data

It provides a lean, flexible, and ultra-scalable deep learning framework that enables users to create state of the art recommendations for complex applications that use time series data. maxML can be applied to applications in healthcare, industrial, telecommunications, finance and other industries that generate time serialized event data from its operations.

maxML is typically deployed in 3 key workflow steps. It can be applied to historical data, but the best applications are streaming solutions for real time recommendations.

Step 1 – Time Series Predictor

The Time Series Predictor predicts the next value in a time series application using deep learning because of its capability to automatically discover complex features without having any domain knowledge.

Step 2 – Anomaly Predictor

The Anomaly Predictor is responsible for tagging the given time series data points as normal or abnormal using the automatic feature learning capability of the maxML deep learning framework.

Step 3 – Recommended Actions

Recommendations on dealing with the anomalies are built on a collaborative filtering recommender based on an algorithm that can use data from a wide variety of application preference indicators.

introduction to Deep Learning

This is a great series of videos by Grant Sanderson of 3blue1brown that explains the essence of Deep Learning in a very simple and understandable way.

Watch Deep Learning videos

GetTing Started

Build your own intelligent Recommender for time series data on the maxML Deep Learning platform. 

maxML is typically deployed in 3 key workflow steps. It can be applied to historical data, but the best applications are streaming solutions for real time recommendations.

Stay Up To Date With Our Newsletter

Stay up to date with the latest on maxML releases and news on deep learning, AI and predictive recommender capabilities