Turi Create simplifies the development of custom machine learning models. You don't have to be a machine learning expert to add recommendations, object detection, image classification, image similarity or activity classification to your app.
- Easy-to-use: Focus on tasks instead of algorithms
- Visual: Built-in, streaming visualizations to explore your data
- Flexible: Supports text, images, audio, video and sensor data
- Fast and Scalable: Work with large datasets on a single machine
- Ready To Deploy: Export models to Core ML for use in iOS, macOS, watchOS, and tvOS apps
As you use Turi Create in your work, reference this guide to understand:
- Data ingestion and cleaning with SFrames (and their data-type-specific equivalents)
- Basics of predictive model development: algorithm- and application-based toolkits
- How to evaluate, visualize, and improve upon your model
If you haven’t already installed Turi Create, you can find instructions here.
SFrame is a scalable, tabular, column-mutable dataframe object. The data in SFrame is stored column-wise, and is stored on persistent storage (e.g. disk) to avoid being constrained by memory size. Each column in an SFrame is a size-immutable SArray, but SFrames are mutable in that columns can be added and subtracted with ease. An SFrame essentially acts as an ordered dict of SArrays.
Currently, we support constructing an SFrame from the following data formats: .csv (comma separated value) file, SFrame directory archive (A directory where an Sframe was saved previously), general text file (with csv parsing options; see read_csv()), Python dictionary, pandas.DataFrame and JSON.
An SFrame can be constructed with data from your local file system, a network file system mounted locally, HDFS, Amazon S3, or HTTP(S).
Turi Create offers a broad set of packaged application based toolkits as well as algorithms for model creation.
Application-oriented toolkits in Turi Create offer default parameters, building blocks and baseline models that help you get started quickly with their dataset without sacrificing the ability to go back and customize models later. Each incorporates automatic feature engineering and model selection.
Using these toolkits, you can tackle a number of common scenarios:
- Recommender systems
- Image classification
- Drawing classification
- Sound classification
- Image similarity
- Object detection
- Style transfer
- Activity classifier
- Text classifier
Essential machine learning models, organized into algorithm-based toolkits:
Refer to the Machine Learning API documentation for complete algorithms and toolkits.
With Core ML, you can integrate machine learning models into your macOS, iOS, watchOS, and tvOS app. Many models created in Turi Create can be exported for use in Core ML.