Home

What is DynaML?

DynaML is a Scala environment for conducting research and education in machine learning. DynaML comes packaged with a powerful library of classes for various predictive models and a Scala REPL where one can not only build custom models but also play around with data work-flows.

DynaML uses the newly minted Wisp plotting library to generate aesthetic charts of common model validation metrics.

Motivation behind DynaML

DynaML was born out of the need to have a performant, extensible and easy to use Machine Learning research environment. Scala was a natural choice for these requirements due to its sprawling data science ecosystem (i.e. Apache Spark), its functional object-oriented duality and its interoperability with the Java Virtual Machine.

Current status/Road ahead

DynaML is a fledgling open source project that is in a phase of rapid expansion. Currently it supports.

• Regularized Ordinary Least Squares
• Logistic and Probit Models for binary classification
• Regression with kernel based Dual LS-SVM
• Regression with Gaussian Processes
• Feed forward Neural Networks
• Committee Models
  • Neural Committee Models
  • Gaussian Process Committee Models
• Model Learning and Optimization
  • Gradient Descent
  • Conjugate Gradient
  • Committee Model Solver
  • Back propogation with momentum
  • LSSVM linear solver
• Model tuning
  • Grid Search
  • Maximum Likelihood (ML-II)
  • Coupled Simulated Annealing
• Model validation metrics (RMSE, Area under ROC)
• Entropy based data subset selection
• Data Pipes for configurable workflows

Going ahead we aim to introduce (but not limit to)

• Sampling based Bayesian models
• Large scale committee models (Apache Spark RDD based implementations)
• GPU support