Home General 5 Ways to Overcome Machine Learning Adoption for Enterprise

5 Ways to Overcome Machine Learning Adoption for Enterprise

There’s still a lot of adoption barriers for people to get over when it comes to artificial intelligence (AI) and machine learning (ML).

According to a recent SAP survey, less than 10% of companies were found to be investing in machine learning.

Many of the reasons as to why businesses still aren’t ready to jump onboard include barriers such as unqualified staff, lack of datasets in which to train algorithms and not yet evolved tools and practices.

To try and encourage firms to overcome their reservations about AI and ML, Deloitte researchers have identified 5 ways in which to make it easier and less expensive to use machine learning in their business models:

1. Reduce the amount of data needed for training

Traditionally, training a machine learning model required a large amount of data in which to be anywhere near effective.

However, a recent tool tested by Deloitte staff offers that same level of accuracy from using just a fifth of the training data.

2. Deploy locally

According to the report, the adoption of ML will rise in places where it can most improve efficiency and overall outcomes.

The advancement of AI makes it much easier to use innovative technology while on mobile devices, including language translation and image recognition.

MORE – Essential Enterprise AI Companies Landscape

3. Explain results

One of the main reasons that people don’t tend to trust AI and ML is that’s it’s often hard to explain how it arrived at its decision.

But, this is getting better and new techniques are being developed that will make these models easier to interpret.

4. Automate data science

Data scientists are in high demand. However, around 80% of the work they do could be automated, including algorithm selection and evaluation, data wrangling, feature engineering and selection, and exploratory data analysis.

“Automating these tasks can make data science not only more productive but more effective,” stated the report.

5.Accelerate training

By using GPUs, field-programmable gate arrays, and application-specific integrated circuits, the time it takes to train these models is cut considerably.

Microsoft proved this when it used GPUs to develop a system that recognized conversational speech and humans in less that 12 months.

“Machine learning has already shown itself to be a very valuable technology in many applications,” stated the report.

“Progress along the five vectors can help overcome some of the obstacles to mainstream adoption.”

Source Deloitte

Subscribe to our newsletter

Signup today for free and be the first to get notified on the latest news and insights on artificial intelligence

KC Cheung
KC Cheung has over 18 years experience in the technology industry including media, payments, and software and has a keen interest in artificial intelligence, machine learning, deep learning, neural networks and its applications in business. Over the years he has worked with some of the leading technology companies, building and growing dynamic teams in a fast moving international environment.
- Advertisment -


AI Model Development isn’t the End; it’s the Beginning

AI model development isn’t the end; it’s the beginning. Like children, successful models need continuous nurturing and monitoring throughout their lifecycle. Parenting is exhilarating and, if...