article thumbnail

Model Interpretability with TCAV (Testing with Concept Activation Vectors)

Domino Data Lab

What if there was a way to quantitatively measure whether your machine learning (ML) model reflects specific domain expertise or potential bias? TCAV “uses directional derivatives to quantify the degree to which a user-defined concept is important to a classification result” ( Kim et al 2018 ). Introduction.

Testing 63
article thumbnail

Will enterprises soon keep their best gen AI use cases under wraps?

CIO Business Intelligence

They had ChatGPT write the script, and other gen AI tools to create a digital person who reads the script, a scalable process with at least one measurable benefit: speed. Helping software developers write and test code Similarly in tech, companies are currently open about some of their use cases, but protective of others.

Insiders

Sign Up for our Newsletter

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

article thumbnail

Interview with: Sankar Narayanan, Chief Practice Officer at Fractal Analytics

Corinium

Fractal’s recommendation is to take an incremental, test and learn approach to analytics to fully demonstrate the program value before making larger capital investments. A properly set framework will ensure quality, timeliness, scalability, consistency, and industrialization in measuring and driving the return on investment.

Insurance 250
article thumbnail

Private cloud makes its comeback, thanks to AI

CIO Business Intelligence

As we are testing and dipping our toes in the water with AI, we are choosing to keep that as private as possible,” he says, noting that the public cloud has the horsepower needed for many LLMs of today but his company has the option of adding GPUs if needed via its privately owned Dell equipment.

IT 143
article thumbnail

What is business analytics? Using data to improve business outcomes

CIO Business Intelligence

Business analytics can help you improve operational efficiency, better understand your customers, project future outcomes, glean insights to aid in decision-making, measure performance, drive growth, discover hidden trends, generate leads, and scale your business in the right direction, according to digital skills training company Simplilearn.

article thumbnail

Agile Technology and Big Data Improve the State of Cybersecurity

Smart Data Collective

The number increased 56% between 2017 and 2018. Cybersecurity experts are using data analytics and AI to identify warning signs that a firewall has been penetrated, conduct risk scoring analyses and perform automated cybersecurity measures. This can be achieved through penetration testing, simulated attacks, and application scanning.

Big Data 110
article thumbnail

MNIST Expanded: 50,000 New Samples Added

Domino Data Lab

Recently, Chhavi Yadav (NYU) and Leon Bottou (Facebook AI Research and NYU) indicated in their paper, “ Cold Case: The Lost MNIST Digits ”, how they reconstructed the MNIST (Modified National Institute of Standards and Technology) dataset and added 50,000 samples to the test set for a total of 60,000 samples. Did they overfit the test set?

Testing 83