ACTICO Machine Learning is a software platform with integrated components for applying machine learning. The platform provides support throughout the process from data preparation to using the trained machine learning models in production systems.
Business decision modeling is the basis for intelligent decision automation. The decision model takes into account all aspects relevant to make the business decision. Some aspects are defined by subject matter experts in the form of business rules, others are learned from data and implemented in the form of machine learning models.
To create a good machine learning model, data scientists need to prepare data and define relevant characteristics (feature engineering). ACTICO Modeler simplifies this often cumbersome step by enabling graphical feature engineering rules models that can be executed and repeated automatically later on.
Training describes the learning process. Data scientists select a machine learning algorithm, such as Gradient Boosting or Deep Learning, that is used to automatically learn from the prepared data. ACTICO offers an easy configuration and monitoring of the training process, and allows automated retraining.
The trained machine learning model is stored and managed in a centralized repository together with business decision and rule models. Users can administer and govern these models depending on their roles and permissions. A variety of governance features helps control, monitor and document activities across the model life cycle.
Execution means that the machine learning model is used for automated decision-making. The machine learning model is orchestrated via decision or rule models. ACTICO Engine ensures the automated execution of the models – either as web services or directly integrated into solutions for high-performance scenarios.
By executing the machine learning model in day-to-day business operations, new data is collected constantly that can be used to improve the model. Adaptive learning allows the automation of the machine learning process. This way, machine learning becomes efficient and continuously unlocks new data knowledge to improve business decisions.
ACTICO Machine Learning relies on the business rules approach to operationalize machine learning. Generated machine learning models are orchestrated together with business rule models and executed automatically – scalable and with high performance.
ACTICO Machine Learning provides easy access to advanced algorithms and approaches for a wide range of applications. The platform relies on powerful technologies to enable better predictions and drive automation.
ACTICO Machine Learning enables the automatic retraining of machine learning models and facilitates their rapid deployment. This way, new data can be leveraged to continuously improve predictive models, and a self-learning system can be realized.
ACTICO offers high agility when applying machine learning. The integrated platform streamlines the process and allows users to run through the entire process with just a few clicks. Companies can realize self-learning decisions and accelerate the time to market.
ACTICO offers a high ease of use and a trouble-free application of machine learning. With graphical decision modeling, rules-based feature engineering, UI-supported training and 1-click deployments, even users with limited technical skills can use artificial intelligence.
ACTICO technology is built for enterprise-scale scenarios that require high scalability and availability. In-memory technology and clustering allow for high performance of both the machine learning process and the automated execution of the resulting models.