Despite the increasing interest in machine learning (ML), many companies are finding it difficult to fully leverage its advantage because of the difficulty in setting it up effectively [Schreck, B et al., 2018]. We consider the typical stages in ML : Data Collection, Data Preparation, Modelling, Training, Evaluation, Parameter Tuning and Prediction [Yufeng Guo, 2007]. Empirical observations point out a gap in the the early stages, highlighting the urgent need of Software and Requirements Engineering research for ML [Nalchigar,S et al., 2019], [Islam, Md Johirul, et al., 2019] in order to help choosing and implementing ML solutions.
When choosing the ML strategy for a specific task, non-expert users and companies can rapidly become lost in questions such as what type of ML to use, what data is available and whether it matches the chosen type, what the right size of the dataset is, what the computation needs are, what the best model for the existing data is, what kind of performance is expected from the resulting system, among many others [Nalchigar,S et al., 2019]. Typically, such questions are answered with the help of a ML expert. We aim however to give an initial idea about such elements, automatically, by asking the user a few simple questions such as: what is the rough type of the task, how much time and other resources (people that can label data, GPUs etc) are available, as well as some simple characteristics of the available data.
Therefore, the idea of this project is to explore ways of automatically matching the constraints and requirements that a beginner would be able to describe with the typical parameters that a ML expert would use in order to assess and identify the most appropriate ML approaches for the task. The aimed expert system is not willing to help building a solution but simply serve as an initial guide for the ML journey. It’s very important to understand the limitation of the intended system. We would consider it successful if it helps 80% of the users get basic initial guidance for their first ML projects. To put it in a different perspective, the system aims to save some days of work for people (professional and students alike) in their search for basic answers, which can have significant impact considering the great number of people and companies starting ML projects . If in addition it also saves hours of GPU usage by preventing common mistakes, the result is even more rewarding.
The project will give the successful applicant an interesting opportunity to learn about state-of-the-art machine learning methods, models and frameworks, which are highly sought-after skills in today’s job market. In addition to working on the system, the successful candidate will also contribute to a publication about the project.
Schreck, B., Kanter, M., Veeramachaneni, K., Vohra, S., Prasad, R.: Getting value from machine learning isn’t about fancier algorithms – it’s about making it easier to use. Harv. Bus. Rev. (2018)
Yufeng Guo, “The 7 Steps of Machine Learning,” 2017, https://towardsdatascience.com/ the-7-steps-of-machine-learning-2877d7e5548e
Amershi, Saleema, et al. "Software engineering for machine learning: a case study." Proceedings of the 41st International Conference on Software Engineering: Software Engineering in Practice (ICSE). IEEE Press, 2019.
Islam, Md Johirul, et al. "What Do Developers Ask About ML Libraries? A Large-scale Study Using Stack Overflow." arXiv preprint arXiv:1906.11940 (2019).
Nalchigar S., Yu E., Obeidi Y., Carbajales S., Green J., Chan A. (2019) Solution Patterns for Machine Learning. In: Giorgini P., Weber B. (eds) Advanced Information Systems Engineering. CAiSE 2019. Lecture Notes in Computer Science, vol 11483. Springer, Cham
NAVER LABS is a world class team of self-motivated and highly engaged researchers, engineers and interface designers collaborating together to create next generation ambient intelligence technology and services that are rich with the organic understanding they have of users, their contexts and situations.
Since 2013 LABS has led NAVER’s innovation in technology through products such as the AI-based translation app ‘Papago’, the omni-tasking web browser ‘Whale’, the virtual AI assistant ‘WAVE’, in-vehicle information entertainment system ‘AWAY’ and M1, the 3D indoor mapping robot.
The team in Europe is multidisciplinary and extremely multicultural specializing in artificial intelligence, machine learning, computer vision, natural language processing, UX and ethnography. We collaborate with many partners in the European scientific community on R&D projects.
NAVER LABS Europe is located in the south east of France in Grenoble. The notoriety of Grenoble comes from its exceptional natural environment and scientific ecosystem with 21,000 jobs in public and private research. It is home to 1 of the 4 French national institutes in AI called MIAI (Multidisciplinary Innovation in Ai) It has a large student community (over 62,000 students) and is a lively and cosmopolitan place, offering a host of leisure opportunities. Grenoble is close to both the Swiss and Italian borders and is the ideal place for skiing, hiking, climbing, hang gliding and all types of mountain sports.