IBM is aiming to use artificial intelligence inside of its Watson Studio data science platform to automate data prep and some of the drudgery needed before rolling out enterprise AI.
The data prep and governance tools are included in AutoAI. AutoAI will be core to Watson Studio and its efforts to free up data scientists to focus on models. According to Forrester, 60% of respondents cited data quality as a big hurdle to deploying AI.
Primers: What is AI? | What is machine learning? | What is deep learning? | What is artificial general intelligence? |
Top cloud providers 2019: AWS, Microsoft Azure, Google Cloud; IBM makes hybrid move; Salesforce dominates SaaS | Enterprise AI and machine learning: Comparing the companies and applications
Watson Studio is combining AutoAI along with its Watson Machine Learning to automate core steps in deploying AI. AutoAI is available in Watson Studio on IBM Cloud. Key features include:
- Automation of data preparation and preprocessing.
- Model development and feature engineering.
- Hyperparameter optimization to build data science and AI models.
- A suite of model types for data science including gradient boosted trees and processes to experiment with machine learning.
- IBM Neural Networks Synthesis (NeuNeuS), which is in open beta. NeuNueS allows data scientists to optimize speed or accuracy and track model training.
For IBM, AutoAI is part of a Watson Studio build out as well as a broader data science portfolio that includes IBM Watson Assistant and Discovery and Watson Machine Learning.
Recent Watson moves:
- IBM updates Watson Studio
- IBM aims to make Watson a set of microservices that can run across multiple clouds
- IBM aims to meld AI with human resources with Watson suite
- IBM launches pretrained Watson packs for industries
- The AI, machine learning, and data science conundrum: Who will manage the algorithms?