Neural network models and datasets of billions of records have been around for a few decades now. It is in recent years though that machine learning has become most popular, and that is mainly due to three factors. First, the amount of data that is being generated everyday and needs to be analysed has increased by several orders of magnitude. Second, the rise of cloud computing has made data storage and processing accessible to most. Third, open source technologies have allowed many developers to build and use machine learning algorithms, instead of just an elite few who use expensive commercial software within large organisations. All these factors have fuelled a renewed interest in machine learning research that has lead to significant breakthroughs in the field.
In this report we explore the core of the new ecosystem that is emerging: data scientists and machine learning developers. Based on the experiences of 2,280 data scientists and machine learning developers from 145 countries we explore how they are involved in machine learning, their goals, the technologies they use, the challenges they face and if and how they attempt to monetise machine learning.
The Machine Learning & Data Science Landscape report comes with a single user license. If you are interested in an enterprise-wide license please contact us.
- Is machine learning mostly the realm of curious amateurs and students or of professionals?
- To what extent are data scientists involved in other closely linked development sectors such as AR/VR, IoT, backend, desktop and mobile?
- What is the balance between building and consuming machine learning algorithms?
- How do different regions compare in their machine learning activities?
- Are image classification and speech recognition mostly researched or productised?
- How popular are recommendation systems, conversational patterns in text, fraud detection and a dozen more application areas?
- How do commercial machine learning toolkits compare to the open-source ones and how many developers prefer to build their own algorithms?
- In which application areas is processing power more of a challenge?
- Which is currently the most lucrative revenue model for machine learning applications?
- Are recommendation systems paying off?
To get a full list of this report’s contents and a sample graph, please download the brochure.
This report is based on the 11th edition of Developer Economics, a large-scale online developer survey designed, produced and carried out by SlashData over a period of six weeks between April and May 2016. The survey received 16,500+ responses from mobile, desktop, IoT, cloud, augmented / virtual reality and machine learning developers, as well as data scientists. The survey received 2,280 responses from data scientists and machine learning developers that formed the basis of the insights of this report.
The online survey received responses from developers across 145 countries and was translated into nine languages (Portuguese, Spanish, French, Turkish, Russian, Korean, Japanese, Chinese (Simplified) and Vietnamese). It was promoted by 59 leading community and media partners within the software development industry.
To minimise regional, platform and other forms of sampling bias, we cleansed and weighted the survey raw data. For more information on our methodology please visit our methodology page https://www.slashdata.co/methodology