When it comes to evaluating data, no one in this dot-com era can bring worthy results except Google. But now, the search giant is on the move to enable visualisation of high-dimensional, bulky data using its Embedding Projector.
Open source Embedding Projector gives an intuitive way of exploring data that is used by machine learning systems. It allows you to views various data patterns in 2D and 3D modes.
“With the Embedding Projector, you can navigate through views of data in either a 2D or 3D mode, zooming, rotating and panning using natural click-and-drag gestures,” the Google team led by Daniel Smilkov wrote in a blog post.
The tool includes three methods of data dimensionality reduction to enable easy visualisation of complex data. These include PCA, t-SNE and custom linear projections.
Standalone TensorFlow
Further, Google has developed a standalone version of its TensorFlow machine learning library to let you easily visualise and high-dimensional data without requiring any installation. You just need to visit the Projector webpage on the TensorFlow website to visualise and examine your data.
The open source availability of Embedding Projector is likely to help Google reach new levels in the world of machine learning. It would also expand the functionality of TensorFlow that has already been competing against Facebook’s Torch framework.
