Failure Model Prediction Using Machine Learning Techniques

Using multivariate time series data, its describes a visual representation of a failure model that may be used to predict when and where machines will break down throughout an
industrial production process. A method for cleaning and pre-processing the data and for training and evaluating a prediction model will be shown. The modular and scalable design of the prediction dashboard model ensures that it can adapt to shifting requirements when manufacturing methods and data sets are updated.