POLYMATICA ANNOUNCES THE RELEASE OF THE POLYMATICA ML MACHINE LEARNING MODULE

Polymatica, the world famous developer of zero-coding analytical systems, has presented the first commercial version of the Polymatica ML module, which enables business users to create machine learning models and control their life cycle.

We are pleased to present the first version of Polymatica ML, a data mining module that uses machine learning methods. It allows users without programming skills to go through the entire cycle of creating and applying models. The cycle consists of the following stages:

1. conduct preliminary data analysis to identify the most significant indicators for building future models;
2. build several models using a simple constructor, compare their quality and determine the champion model;
3. select the method of applying the model on the main data set – there are several of them, depending on the problem being solved.

For effective industrial operation, there is a built-in mechanism for monitoring the model, since over time its accuracy decreases and an adjustment is required.

Polymatica ML 1.0 Features

Data Discovery:

• Connecting to various data sources and querying data.
• Exploratory data analysis.
• Calculation and analysis of different statistical characteristics of datasets.
• Detection of patterns and dependencies in data.
• Assessment of suitability of data for modeling.
• Selection and transformation of data for modeling.

Model Designer:

• Building machine learning models in visual interface.
• Validation of models.
• Interpretation of models.
• Assessment of models’ quality.
• Comparison and selection best models.
• Ability to add different Python ML libraries through system configuration.

Model Manager:

• Centralized repository for models built in Polymatica ML as well as external models built in Python.
• Model lifecycle management with the flexible configuration of stages, roles and workflows.
• Deployment of models to different targets: online, batch, interactive.
• Monitoring and evaluation of models’ accuracy and performance.
• Retraining models on new data.

“When we started developing the module, we chose two goals: democratizing machine learning for business users and creating an easy-to-use lifecycle management platform to keep all your models in the state of high efficiency,” comments Konstantin Malashenko, Product Director for Polymatica ML at Polymatica. “In a short time, we were able to assemble a first-class team with experience in international companies and create a product that is a constructor that allows us to assemble machine learning models from ready-made components, which significantly reduces the time from formulating a hypothesis to testing it. In doing so, we went beyond just creating models, targeting the ModelOps niche or enterprise-level machine learning model management. This allows you to store all models in a single repository, systematically monitor the quality of their work and reuse existing algorithms in new models. In the first version, this functionality is already available, and we plan to actively develop it in the near future”.

The practical implementation of Polymatica ML is already gaining momentum: the module is involved in several pilot projects in different industries: predicting the response to advertising campaigns for retail, predicting breakdowns of heat supply pipes for the power industry, monitoring employee health in hazardous industries.