DataTalks # 4: Predictive Analytics

    On October 10, the fourth DataTalks took place. This time the topic of the meeting was predictive analytics, and we would like to share video recordings of speeches with the community.

    Why predictive analytics?
    It allows predicting various future events based on historical data, such as customer behavior or the results of actions taken. Thanks to it, a business can make optimal decisions and take into account forecasts of future actions and the wishes of their customers.

    Under the cut you will find recordings of performances:
    • How to answer the question “What will happen?”: Practical tips / Andrey Yarmola, Data Science Team Lead at Wargaming
    • Necessary minimum tools for building your recommendation system / Alexei Dyomin, Java Server Side Developer at InData Labs
    • What is predictive analytics and who needs it / Nadezhda Rucanova, deputy director of the representative office of SAP CIS LLC, Mikhail Avetisov, leading expert of predictive analytics LLC SAP CIS
    • Building a data warehouse based on the Hadoop platform / Igor Nahvat in Data Integration Engineer, Wargaming
    • Using predictive analytics to manage client base value / Maxim Mozgovoy Director of CRM at Wargaming

    image



    How to answer the question “What will happen?”: Practical advice / Andrey Yarmola
    Data Science Team Lead, Wargaming
    During his professional career, Andrey managed to apply his knowledge of data analysis in science and business. In his report, he shared his experience and told how to make forecasts prepared using special techniques correctly understood and taken for granted by stakeholders.
    The report will be of interest to analysts, as well as to those who make decisions based on forecasts.



    Necessary minimum tools for building your own recommendation system / Alexey Dyomin
    Java Server Side Developer, InData Labs
    More and more companies are realizing the importance and value of personalized user interaction. In the report, Alexey gave several examples of simple solutions that will allow you to build a service for creating personalized recommendations with minimal time costs.
    The report will be useful to technical specialists answering queries like: "we want to start at least with something, but we don’t know which way to approach."



    What is predictive analytics and who needs it / Nadezhda Rucanova, Mikhail Avetisov
    Deputy Director of the representative office of SAP CIS LLC and leading expert of SAP CIS LLC on predictive analytics
    From the report of Nadezhda and Mikhail you will learn about current achievements in the development of machine learning theory, and You will also see how predictive analytics work with specific examples. All calculations were carried out using the SAP software online.



    Building a data warehouse based on the Hadoop platform / Igor Capture
    Data Integration Engineer, Wargaming
    Igor talked about the architecture of the Wargaming data warehouse, based on technologies from Cloudera and Oracle. We are talking about both the source data sources and the experience of their integration using independently developed solutions and specialized tools, such as Apache NiFi.
    The presentation will be interesting for technical specialists who are already familiar with the Hadoop stack.



    Using predictive analytics to manage customer base value / Maxim Mozgovoy
    Director of CRM, Wargaming
    Happy and satisfied regular customers are the basis of prosperity for many companies. However, the outflow of customers is rarely the subject of close attention - usually we see a decrease in income. And we think, accordingly, about its increase, and not about retaining users and managing the client base. This leads to the fact that many companies are forced to spend resources on attracting one person several times.
    In his speech, Maxim talked about what a client base is, what is its value, and how to use it as the basis of a client strategy. Maxim also spoke about KPI metrics and systems for solving value management problems.

    Also popular now: