New in Wolfram Finance Platform 2.4 Based on the Wolfram Language 11.1, Wolfram Finance Platform 2.4 features enhanced capabilities in key areas, including machine learning, statistics, computation and data visualization. Improved Machine Learning Wolfram Finance Platform 2.4 extends and improves its machine learning capabilities. New functions make it easy to extract features, reduce dimension, cluster data, optimize hyperparameters and obtain interpretable models. CUDA accelerated training makes machine learning on large datasets practical. Documentation: Machine Learning Time Series & Signal Processing Extended functionality includes the ability to allow for natural specifications such as monthly duration when applying a function over windows, as well as precise control over the resulting value's placement in time. Date and quantity support is more complete and natural. Documentation: Signal Processing Documentation: Time Series Documentation: Time Series Processes Statistics With new statistical distributions, support for money units throughout the system, enhancements across hypothesis testing, parameter estimation and sampling, the Wolfram statistics framework is stronger, faster and easier to use. Enhanced Visualization New visualizations for multidimensional data, textual data and timelines are supported by better labeling and plotting over nonrectangular regions, making it easier than ever to create compelling and insightful plots. Documentation: Labeling Documentation: Exclusions Enhanced Mathematical Computation Numerical PDE-solving capabilities now include events, sensitivity computation, new boundary condition types and improved complex-value PDE solutions. Documentation: Partial Differential Equations/a> Text Processing Tools for handling multiple languages and character sets enable advanced processing and analysis of structured, unstructured and semantic text. Documentation: Text Analysis Knowledgebase The Wolfram Knowledgebase provides faster, more convenient access to the world's computable data, which can now be augmented with custom entities and your own data. New datasets, such as Federal Reserve Economic data, are also included. Documentation: Knowledge Representation and Access Documentation: EntityStore » Network Communications New Wolfram Cloud communication protocols enable a more abstract and scalable communication system to facilitate asynchronous communication between sessions. Documentation: Channel-Based Communication
