DataRobot has gained traction in the AutoML world due to its intuitive platform that can be leveraged to build ML models without the need for data scientists. In an attempt to further enhance the platform, the firm introduced Visual AI in the DataRobot 6.0 to automate ML workflows with computer vision technology.

The company has been making great strides in the data science landscape. It has been committed to continually improve its platform to simplify the workload of data scientists and in turn, bring efficiency within organizations. In December 2019, DataRobot acquired Paxata to enhance its platform’s capabilities after bagging $200 Million in Series E funding. The firm has been aggressively acquiring and integrating new features to streamline AI workflows.

What is Visual AI  

Computer vision technology has become the foundation for many AI-based applications, such as facial recognition, and object detection, among others. Consequently, DataRobot has come up with a new solution to simplify the incorporation of image data into ML models alongside tabular and text-based data types.

With this, anyone can build models by just dragging and dropping images into the DataRobot platform. One can also get started with the models by feeding only a few hundred images for training, thereby getting outputs within minutes or hours. Also, you do not require GPUs as it is optimized to perform even on basic hardware. This is because the firm already offers pre-trained neural networks that do the heavy lifting.

How Is It Different?

The Visual AI solution empowers users to build binary and multiclass classification and regression models with images. One can use it to develop completely new image-based models. Besides, users can also add pictures as new features to existing models to enhance their accuracy from totally different images. In other words, the variables can be extended even after the model is trained, resulting in more flexibility in the data science workflows. 

Getting started with the solution:

These solutions are self-explanatory and point to the factors that lead to the outcome of models. The steps are as follows:-

1. Create a zip file of images: Keep images into different folders or use comma separated values (CSV) file if you want to include additional features; and then zip the files.
2. Drag and drop the zip file into the platform: Upload or drag and drop the zip file into a new project and pick your target.
3. Explore your data: The platform will perform exploratory data analysis (EDA) and show interesting statistics, along with missing and duplicate data.
4. Model training: It will automatically train, test, and compare different models to recommend the best one eventually.
5. Evaluation: The models can be evaluated for its accuracy using automated visualizations or neural network visualizer, image embeddings and activation maps. The solution shows how the model pre-processed the data, and why a particular algorithm was picked over the other, and where the neural network looked in the image for every single prediction.
6. Tune and tweak: If needed, users can tune the models by changing the hyperparameters. However, an expert is required to ensure proper values selection for hyperparameters. 
7. Deploy and monitor: On deploying the model into production, the solutions provide a view where one can monitor, manage, and enhance the performance. 

Performance Of The Solution

The solution was tested with ~14k images of natural scenes such as buildings, forests, streets, mountains, etc. Visual AI trained 40 models in just two hours without the support of GPUs. The accuracy was around 92%, which was tuned to get even higher results.

Although the results were exceptional, one may have to check for bias and approach an expert to interpret the explainability. Consequently, one cannot rely on the solution even though the models are transparent. AutoML has been helping data scientists, but it will be strenuous for non-experts to work with such tools.

Outlook 

Nevertheless, Visual AI delivered results which were higher than one would have envisioned, but the firm cannot guarantee similar results unless it is tested with different types of data. It is undoubtedly a great addition to DataRobot offerings, but might still require experts to make the most of it. But this has been the case with other AutoML solutions – these still need specialists to enhance outputs. One cannot directly use outputs of models from AutoML solutions and make business decisions based on it, especially when experts are critical of the computer vision technology due to its potential bias outcomes.

The post HOW TO GET STARTED WITH VISUAL AI – THE NEW AUTOML SOLUTION BY DATAROBOT appeared first on Artificial Intelligence.

DataRobot, the leader in enterprise AI, announced enhancements to its enterprise AI platform, including AI Applications, Automated Deep Learning, and Visual AI. These new introductions further unlock the value of AI by putting the power of AI into the hands of more users and making it simpler to build and deploy deep learning models.

“Subject matter experts from any industry can now solve new business problems by including relevant image-based content along with other more traditional data types. This latest evolution of our platform will empower users to leverage AI to make even better decisions based on broader perspectives.”

In the latest version of the platform, DataRobot has introduced:

• Visual AI: With Visual AI, users can address computer vision use cases and combine incredibly diverse types of data in their models. Visual AI offers immediate support for use cases requiring image recognition and classification. Users can simply drag and drop a collection of images into a project and build custom deep learning models in minutes. DataRobot’s Visual AI then takes image-based machine learning one step further by allowing users to leverage images alongside any other feature types such as numeric, categorical, dates, and raw text.
• AI Applications: With the latest platform release, any machine learning model, including DataRobot-generated models or models written in R or Python, can be turned into an AI application. This enables employees of all skill levels to interact with the predictive insight of the underlying model and experiment with different scenarios, predict results, and make more informed business decisions. The new feature also includes an Applications Gallery – a one-stop shop that allows business users to find the application that best suits their needs.
• Automated Deep Learning: DataRobot has significantly boosted its deep learning capabilities, powered by a new Keras-based model framework for which DataRobot recently secured a provisional patent. Traditionally, training deep learning models is expensive and time consuming. DataRobot’s new capabilities allow users to build successful and reliable deep learning models that are ready to deploy into production. The new capabilities also make it easy to understand these models – all with the infrastructure a user has in place.

“Having pioneered the automated machine learning category, we are proud to push the boundaries of what’s possible with the technology by offering these novel automated deep learning and Visual AI capabilities,” said Phil Gurbacki, SVP of Product and Customer Experience, DataRobot. “Subject matter experts from any industry can now solve new business problems by including relevant image-based content along with other more traditional data types. This latest evolution of our platform will empower users to leverage AI to make even better decisions based on broader perspectives.”

Additionally, DataRobot has unveiled enhancements to:

• MLOps: In this release, DataRobot MLOps has been enhanced to include pre-packaged model environments so users can drag-and-drop model files, developed in languages such as Python and R, and deploy them using Kubernetes. The release also includes unlimited batch scoring with integrations to leading cloud storage options for massive scale. Lastly, the enhanced MLOps solution offers Monitoring Agents that can capture metrics from models deployed to almost any environment.
• Automated Time Series: Automated Time Series now features new deep learning techniques that remove the traditional forecasting barriers to make easy work of large-scale multi-series forecasting applications.
• DataRobot Paxata: Following the acquisition of Paxata in December 2019, DataRobot has integrated Paxata’s AI-assisted data preparation solution seamlessly with its AI Catalog to empower novice and expert users to rapidly explore, clean, combine, and shape data for training and deploying machine learning models.

The post DataRobot Unveils Latest Version of Enterprise AI Platform, Introducing Visual AI, AI Applications, and Automated Deep Learning appeared first on Artificial Intelligence.