Hitachi Vantara has announced new capabilities have been added to the platform to enhance machine learning orchestration in production.
The new capabilities will help data scientists monitor, test and redeploy models in continual production much faster and efficiently than they currently do.
Collectively known as a ‘Machine Learning Model Management’ the capability uses the new tools within a data pipeline to help improve business outcomes and reduce risks. This is done by making the process much easier to update models that are in continual change.
This process currently requires a significant amount of manual work, which is not done frequently enough leading to mistakes being made and lower efficiency. Research carried out by Hitachi Vantara found that two thirds of organisations do not have an automated process to seamlessly update their predictive analytics models and as a result less than one quarter of machine learning models are updated daily.
The new capabilities will improve efficiency and quality.
Key features to the new capabilities include getting models into production much quicker, using machine learning orchestration steps that evaluate models and improve accuracy using real production data prior to going live. Secondly, the capabilities will maximise the model accuracy whilst in production using evaluation statistics to help identify degraded models. This will make it easier to analyse model performance and uncover errors.
The final element of the new capabilities is collaborating and governing model operations at a larger scale. The new capabilities promote collaboration within operating teams, data scientists, developers and app architects to provide greater transparency. As a result, this allows data and data pipelines to easily be shared across the workforce, standardises and reused to allow machine learning applications to be built faster.
Google offers AI, ML courses for free to the public
Hitachi Vantara’s IoT, data driven ethos
MWC demonstrated a vast amount of the emerging technology
“Machine learning and artificial intelligence (AI) are optimising everything from customer interactions to enterprise operations. As these applications evolve, data scientist and IT operation teams will need to move newly trained models into production faster than ever before, which can jeopardise model accuracy, collaboration and governance,” said John Magee, VP, product marketing, Hitachi Vantara.
Content from our partners
“Hitachi Vantara Labs’ machine learning model management provides improved algorithmic transparency and automation so application teams can focus their efforts on innovating rapidly without risking model deterioration.”
The model and new capabilities are now available for access in the Pentaho Marketplace.
