Vitaliy Kononenko

CTO at Computools

According to the Gartner curve, Machine learning is in the process of transitioning from a hype climb and high expectations to a productivity plateau, and AutoML is at the peak of expectations. These 2 directions will be popular and most in-demand within 2-5 years.

Statistical Observations:

• According to the forecasts of International Data Corporation, by 2021, the costs of the AI ​​and ML industry will amount to $ 57.6 billion, while in 2017 amounted to $ 12 billion.
• Automated Machine Learning (or AutoML) has been gaining popularity over the last couple of years. Several companies have already launched products that demonstrate the potential pf this service.
• According to the LinkedIn report on new jobs for 2018, the number of jobs in the ML area has grown 12 times in the last calendar year.
• According to Forbes, by 2020, artificial intelligence (AI) and machine learning (ML) should create a total cost of $ 2.6T in the field of marketing and technology of the new century, as well as an additional $ 2T in the field of production and supply chain planning.

The list of trends that will remain relevant:

1. Providing visualizations and insights to help describe and understand the data.
2. Search, build, extract the best features for a given dataset.
3. Building more powerful, more intelligent predictive models.
4. Bridging the gap between black-box modelling and the use of such a model.
5. Facilitate the production of these models.

All data for the study were obtained from open sources.