PLATEFORME DE MACHINE LEARNING

Our Machine Learning platform can process large volumes of industrial data within seconds. We currently offer an anomaly detection algorithm that can be used by any operator in a factory (it’s SO simple and therefore SO impressive). Later this year, we plan to release an optimization algorithm to help workshop teams identify the best settings to maximize production, energy efficiency, and quality. Our platform leverages two types of algorithms: Anomaly Detection and Optimization. For anomaly detection, we use: Isolation Forests: An anomaly detection algorithm used to identify outliers in a dataset. Matrix Profile: This consists of two main components: a distance profile and a profile index. To detect regime shifts, we primarily use: a) EWMA (Exponentially Weighted Moving Average): A method to identify significant changes in a data series. b) PELT Algorithm (Pruned Exact Linear Time): A technique for detecting change points in time series data. c) Bayesian Online Change Point Detection: A novel approach based on Bayesian theory to identify change points. For process optimization, we typically rely on Probabilistic Graphical Models: Statistical models that encode complex multivariate probability distributions using graphs. What makes our solution groundbreaking: Anyone with basic internet access and very limited knowledge can use our system. We've worked and innovated tirelessly to create an automated, user-friendly interface using the latest technologies.