-MBD
Teko-MBD is a comprehensive Machine Learning (ML) library that is part of the Apache-Spark platform. It provides a variety of algorithms, models, and tools for data science, machine learning, and artificial intelligence applications. Its versatile nature makes it suitable for a variety of user tasks and data types, while its modularity allows users to choose the tools they need most. The primary components of Teko-MBD are Data Science, Machine Learning, Natural Language Processing, and Deep Learning.
The Data Science component of Teko-MBD is focused on providing statistical models for extracting meaningful insights from data. This component enables users to explore, visualize, analyze and interpret their data sets, as well as perform predictive analysis.
The Machine Learning component of Teko-MBD enables users to train and deploy models built on different types of data. It includes a wide variety of algorithms, ranging from classic Supervised Learning algorithms such as Support Vector Machines to more recent unsupervised learning models such as Clustering. It also includes libraries for working with Big Data and GPU-accelerated learning.
The Natural Language Processing component of Teko-MBD provides a variety of tools to analyze text data in order to extract meaningful insights. It includes text cleaning, stemming, sentiment analysis, topic modeling, and other related algorithms.
The Deep Learning component of Teko-MBD provides users with access to powerful, GPU-accelerated neural networks. This component of the library enables users to quickly design and train models with minimal effort. It also provides many pre-trained models for users to reuse and customize.