Deep learning challenges: Lessons from the field

The remarkable momentum that deep learning has achieved is a result of increased computing power, the availability of large datasets and recent advances in artificial neural networks. Increasingly, it is not only the large technology companies, but also smaller companies in other areas that have identified the potential, not just of traditional machine learning, but specifically of deep learning, to make them competitive and successful.

The true potential for AI will only become a reality when more organizations embed AI into their standard operating model

Why read this

This whitepaper is for business stakeholders wishing to gain a sound understanding of the key challenges between conventional software engineering projects and those involving machine learning or deep learning. The paper will highlight 12 potential challenges that may affect AI projects within an organization.

In this paper, case studies are used to illustrate the potential of deep learning as a mainstream tool for a spectrum of businesses. The case studies describe the experiences of Peltarion’s data scientists working in the field on customer projects. As real-life examples, they are used to illustrate the unique challenges of deep learning projects. In sharing them, we seek to use our prior experience to enable businesses to take advantage of deep learning technologies without falling victim to project pitfalls. Deep learning is still a nascent technology and organizations seeking to use it should be informed of these challenges in advance.

This paper is based on a recent academic publication, authored by some of the Peltarion team, and extended for a business audience.

Get your copy here

Download as PDF.

Try the platform