In-company Data Science training and thematic mitap in Voronezh

    On May 25, the second Metaconf meeting will be held in Voronezh , this time dedicated to machine learning. There are five reports in the mitap program, free registration is available here . In particular, Anton Dolgikh, DataArt expert on AI-projects in the field of healthcare, will talk about the “Neural Network Probabilistic Model of the Natural Language”. Today we asked Anton to talk about the experience of systematizing machine learning knowledge within DataArt.

    The scope of ML is constantly expanding (from healthcare to the travel industry). Inside DataArt, at some point, the number of ML development requests exceeded a critical value. Prior to this, we were able to solve such problems by the engineers who worked in the company.

    When it became difficult to manage with our own resources, two development paths became apparent: to hire new employees or train specialists within the company. In the first case, we face the risk that the ML developer we hired afterwards does not immediately fall into a new project from our professional field. At the same time, people who are narrowly engaged in machine learning are usually not ready to engage, for example, in fullstack development. Therefore, we relied on DataArt engineers who are interested in developing towards ML, but who are able to return to their previous work if necessary.

    The preparation process needs to be systematized. It may seem that the Internet is filled with tons of online and video courses. But in order to develop productively, a person needs a development vector - from randomly listening to any courses, it is of little use.

    What have we done:

    1. First of all, they formed the core - an initiative group of colleagues with the most experience and expertise in various areas of machine learning. They prepared a series of presentations, made an overview of existing courses and made recommendations: which courses you need to take in order to acquire skills relevant to the tasks that DataArt solves.
    2. We organized math courses. Obviously, ML is inherently mathematical statistics and optimization methods. To understand and correctly use machine learning methods, certain mathematical knowledge is needed. At first glance, specialists who have received a technical education always know mathematics well. But in practice, it turns out that skills are forgotten very quickly. This imposes a restriction on the course: a company, unlike a university, cannot provide fundamental knowledge, but knowledge must be adequate and deep enough. We invited a teacher from the outside to read the course (our colleagues were too busy). The program was focused on areas directly related to machine learning: linear algebra, analysis, probability theory, optimization methods.
    3. Every month, our ML specialists conduct educational seminars on the latest achievements in this field. Recording of seminars is available to all employees of the company.
    4. In addition to seminars, DataArt ML specialists regularly publish a digest of interesting materials (methods, articles, books) with brief annotations and comments.

    The company supports these initiatives, a budget is allocated for the purchase of literature and the participation of colleagues in conferences, for iron and mentoring programs. The result of individual mentoring training is a ready-made prototype that can be used at conferences or at meetings with potential customers. As an example, we can cite the result of the work of our expert Andrei Sorokin - a model that detects and classifies skin lesions ( ). To optimize the resulting model for use on mobile devices, the employee just helped in the framework of the mentoring program. The model took 12th place in the international competition ISIC 2018 , beating not only individual participants, but also university teams.

    The above systematization of the process allowed us to quickly and expertly process all requests from the field of machine learning that come to DataArt from potential customers. We have prepared marketing materials, and sales teams are always available experts who can answer customer questions. Several projects have already been successfully completed.

    Like many large technology companies, DataArt scales expertise and educational programs to an external audience. On May 25, Voronezh hosts an open Machine Learning meetup , the participants of which learn about trends in ML technologies, problems and tasks that can be solved with their help, about real-life projects that use machine learning methods and artificial intelligence.

    Also popular now: