1
00:00:00,390 --> 00:00:05,447
[MUSIC]

2
00:00:05,447 --> 00:00:07,250
Hello, and welcome.

3
00:00:07,250 --> 00:00:08,170
My name is Dimitri, and

4
00:00:08,170 --> 00:00:12,060
I'm happy to see you are interested
in competitive data science.

5
00:00:12,060 --> 00:00:14,710
Data science is all about
machine learning applications.

6
00:00:14,710 --> 00:00:17,640
And in data science, like everywhere else,
people are looking for

7
00:00:17,640 --> 00:00:20,380
the very best solutions to their problems.

8
00:00:20,380 --> 00:00:24,240
They're looking for the models that
have the best predictive capabilities,

9
00:00:24,240 --> 00:00:27,850
the models that make as
few mistakes as possible.

10
00:00:27,850 --> 00:00:31,830
And the competition for one becomes
an essential way to find such solutions.

11
00:00:31,830 --> 00:00:35,340
Competing for the prize,
participants push through the limits,

12
00:00:35,340 --> 00:00:37,680
come up with novel ideas.

13
00:00:37,680 --> 00:00:41,250
Companies organize data science
competitions to get top quality models for

14
00:00:41,250 --> 00:00:42,940
not so high price.

15
00:00:42,940 --> 00:00:45,980
And for data scientists,
competitions become a truly unique

16
00:00:45,980 --> 00:00:49,215
opportunity to learn, well,
and of course win a prize.

17
00:00:50,360 --> 00:00:54,090
This course is a chance for you to catch
up on the trends in competitive data

18
00:00:54,090 --> 00:00:58,060
science and learn what we,
competition addicts and at the same time,

19
00:00:58,060 --> 00:01:01,239
lecturers of this course,
have already learned while competing.

20
00:01:02,390 --> 00:01:05,976
In this course, we will go through
competition solving process step by

21
00:01:05,976 --> 00:01:09,982
step and tell you about exploratory data
analysis, basic and advanced feature

22
00:01:09,982 --> 00:01:13,712
generation and preprocessing,
various model validation techniques.

23
00:01:13,712 --> 00:01:18,498
Data leakages, competition's metric
optimization, model ensembling,

24
00:01:18,498 --> 00:01:20,370
and hyperparameter tuning.

25
00:01:20,370 --> 00:01:25,050
We've put together all our experience and
created this course for you.

26
00:01:25,050 --> 00:01:26,520
We've also designed quizzes and

27
00:01:26,520 --> 00:01:31,000
programming assignments to let you
apply your newly acquired skills.

28
00:01:31,000 --> 00:01:34,570
Moreover, as a final project, you will
have an opportunity to compete with

29
00:01:34,570 --> 00:01:37,545
other students and
participate in a special competition,

30
00:01:37,545 --> 00:01:43,460
hosted on the world's largest platform for
data science challenges called Kaggle.

31
00:01:43,460 --> 00:01:46,354
Now, let's meet other lecturers and
get started.