Developers Club geek daily blog

Big data, Beeline and kokoko

3 years ago
Couple of days ago, having accidentally come on Habr without adblok, I have seen banner:" Beeline, be the man — solve the task shaitan". Chelendzh sounded interestingly, to determine age by set of such parameters, as the region, tariff plan, etc.


Read more »


Mastering of the specialty Data Science on Coursera: personal experience (ch.2)

3 years ago


We publish the second part of post of Vladimir of Podolsk vpodolskiy, the analyst in department on work with formation of IBS which has finished training on specialization of Data Science on Coursera. It is set from 9 kurserovsky courses from Johns Hopkins University + the thesis which successful completion grants the right for the certificate.

Read in the first part: About the specialty Data Science in general. Courses: Instruments of data analysis (programming on R); Preprocessing of data; Documentation of processing of data.

Part 2

Read more »


Moscow schools. As we participated in the second hakaton according to open data

3 years ago

Read more »


Greenplum DB

3 years ago
We continue cycle of articles about the technologies which are used in work of the data storage (Data Warehouse, DWH) our bank. In this article I will try to tell briefly and a little superficially about Greenplum — to the DBMS based on postgreSQL, and which is kernel of our DWH. Will not be provided in article installation log, configs and other — and without it the note has turned out rather volume. Instead I will tell about the general architecture of the DBMS, ways of storage and filling of data, backups, and also I will list some problems which we had faced during operation.



It is a little about our installations:

  • the project lives at us slightly more than two years;
  • 4 circuits from 10 to 26 machines;
  • DB size about 30 Tb;
  • in DB about 10000 tables;
  • to 700 queries per second.

How it works, I ask under kat!

Read more »


Hakaton Big Data for Business: begin the technological startup

3 years ago

We invite developers, analysts, marketing specialists, designers, product managers and business angels on hakaton Big Data for Business – two-day team competition in development of the software products solving business problems through data analysis. Hakaton will pass on November 18-19 in the Kazan IT park. Sponsors of action — the EMC and Brocade company. Partners — Textocat, DGL, Provectus and Business incubator of IT park Kazan. Prize fund — 150 000 rubles.

Having taken part in the hakatena of Big Data for Business, you will be able:

  • to find team of adherents,
  • to think up cool business idea, to implement and improve it with leading experts,
  • to gain recognition,
  • to win valuable prizes,
  • to adopt experience in the technology sphere and the principles of packaging of product,
  • to take the first step towards the startup on the basis of technologies of data analysis
  • to get acquainted with perspective product teams in the field of Big Data.

Further we will tell about key features of our action.

Read more »


Mastering of the specialty Data Science on Coursera: personal experience (p.1)

3 years ago


Recently Vladimir Podolsk vpodolskiy, the analyst in department on work with formation of IBS, has finished training on specialization of Data Science on Coursera. It is set from 9 kurserovsky courses from Johns Hopkins University + the thesis which successful completion grants the right for the certificate. For our blog on Habré it has written detailed post about the study. For convenience we have broken it into 2 parts. Let's add that Vladimir became also the editor of the project on transfer of specialization of Data Science into Russian which have started IBS and ABBYY LS in the spring.

Part 1. About the specialty Data Science in general. Courses: Instruments of data analysis (programming on R); Preprocessing of data; Documentation of processing of data.

Hi, Habr!


Not so long ago my 7-month marathon on specialization mastering "Data science" (Data Science) on Coursera has ended. The organizational parties of mastering of specialty are very precisely described here. In the post I will share impressions of content of courses. I hope, after reading of this note everyone will be able to draw for himself conclusions on, whether it is worth spending time for knowledge acquisition on analytics of data or not.

Read more »


We invite to Media Hack Weekend. On October 16-18, Kiev

3 years ago
On October 16-18 within the Future Media Lab project the most large-scale will take place hakaton in the territory of Ukraine – Media Hack Weekend. About 400 talented specialists, experts and businessmen will gather to solve as media of the future will look. Among the directions of hakaton: big data, virtual and augmented reality, electronic commerce, transmedia storitelling, design and applications programming, creation of game content and many other things.
Experts from the Intel company will also be present on action. They will answer your questions on the Intel technologies, will give consultation on further development of your product. Use the chance to begin the successful project with competent advice of specialists of Intel!
To be registered on Hakaton.

Read more »


Trends of world e-commerce of the market in 2015-2016

3 years ago
The market of electronic commerce both in the world and in RuNet very actively develops, despite crises and other negative phenomena. In the world in year average growth rates according to eMarketer make about 18-20% a year, in Russia and Ukraine growth rates reach 17-18%. These are about 3-4% of the general retail in Russia (in Ukraine slightly less, the market is developed less and now deep crisis) and to 10-12% in the USA and other developed countries. The average level in the world makes about 6%. The only exception, last year in Ukraine because of deep economic crisis the market has not grown in dollar equivalent, but for the local companies it is chance to overtake missed earlier. The most interesting that all of us still are in stage of origin of the market. According to many forecasts the share of electronic commerce in the general retail will reach 20% in the next some years. For the companies of this sector ignoring of this market is equivalent to death today tomorrow.



It is interesting as well that fact that many largest players of electronic commerce in the USA have offline roots, and the USA are certain litmus piece of paper, the catalyst of the market which shows us that will occur in our markets in the same segment in 3-5 years. To it already there is confirmation: in Russia number of large online stores belong large offline networks for a long time and absorption proceed. In Ukraine with it it is more difficult though it is process actively goes, so several months ago the Fokstrot company has redeemed 100% of Sokol.ua online store

Mobile commerce

Read more »


We look for stability in retail, the XYZ analysis of the range

3 years ago
The XYZ analysis — one of forms of the analysis of the commodity range of shop, network or separate commodity group in retail.



The XYZ analysis defines stability of sales of goods for certain period. It is useful to management of the range and deliveries of goods, the organization of work with suppliers. Results allow to separate goods on categories and to select for them place in warehouse, stock rate and the organization of delivery.

As the separate method of the analysis in retail of XYZ is used not so often, more often he can be met as combined with ABC analysis.
But, anyway, as the method for decision-making on management of the range of commodity group or shop can bring undoubted benefit.

Let's begin with consideration of its features and opportunities of application.

Read more »


Two problems of HeadHunter on Data Science Week: try to solve

3 years ago
At the end of August after series of free lectures on Data Science Week 2015, dataton (datathon) – competition where programming teams and analysts were solved by business challenges from the Data Science area.

On the datatena there were three tasks two of which has prepared the HeadHunter team and one company OZON. It was, at once I will tell, not the simplest task because the most part of our data is confidential. Nobody will want that programmers and analysts practised on real summaries or the closed data on vacancies. But all of us have collected something. For check of results organizers have thought up metrics and cheker have written. And these children have won on the datatena:



Directly here and now I suggest you to test the strength and to solve three problems with which children fought on the datatena. Chekera for check and all files I apply.

Read more »