90-95% - 9, 80-89% - 8, 75-79% - 7, 65-74%
- 6, 55-64% - 5, 45-54% - 4, 35-44% - 3, 25-34% - 2, 0-24% - 1You can work in pairs
Best solutions could be presented in class (5 minute talk) to get some extra points
Разделы презентаций
- Разное
- Английский язык
- Астрономия
- Алгебра
- Биология
- География
- Геометрия
- Детские презентации
- Информатика
- История
- Литература
- Математика
- Медицина
- Менеджмент
- Музыка
- МХК
- Немецкий язык
- ОБЖ
- Обществознание
- Окружающий мир
- Педагогика
- Русский язык
- Технология
- Физика
- Философия
- Химия
- Шаблоны, картинки для презентаций
- Экология
- Экономика
- Юриспруденция
Seminar 1 Introduction to Data Science
Содержание
- 1. Seminar 1 Introduction to Data Science
- 2. Grades50% - home assignments, 50% - group
- 3. DefinitionData analysis is the process of transforming
- 4. Data analysis techniquesData mining automatic discovery of
- 5. Two cultures of data analysisData is generated
- 6. Data modeling cultureStarts with assuming a data
- 7. Algorithmic modeling cultureConsiders the inside of the
- 8. Why do you need to learn data
- 9. Data manipulation by Tim Cook https://www.statschat.org.nz/2013/09/11/cumulative-totals-tend-to-increase/
- 10. Even academic superstars may be wronghttp://theconversation.com/the-reinhart-rogoff-error-or-how-not-to-excel-at-economics-13646
- 11. A lot of fraud in science (especially in social sciences)https://www.financial-math.org/blog/2015/10/is-research-in-finance-and-economics-reproducible/
- 12. Random chance plays a huge role in social sciences http://www.tylervigen.com/spurious-correlations
- 13. Intuition might be wrongSimpson’s paradox: graduate admissions to UCB
- 14. Intuition might be wrongSimpson’s paradox: graduate admissions to UCB
- 15. Intuition might be wrong, part 2Monty Hall problemhttps://en.wikipedia.org/wiki/Monty_Hall_problemHumans vs birds: birds win (Herbranson, 2010)
- 16. RR is a language of statistical computingModern
- 17. P.S.Calling Bullshit is a highly recommended online course at the University of Washington http://callingbullshit.org/syllabus.html#Introduction
- 18. Скачать презентанцию
Grades50% - home assignments, 50% - group project96-100% - 10, 90-95% - 9, 80-89% - 8, 75-79% - 7, 65-74% - 6, 55-64% - 5, 45-54% - 4, 35-44% - 3,
Слайды и текст этой презентации
Слайд 2Grades
50% - home assignments, 50% - group project
96-100% - 10,
You can work in pairs
Best solutions could be presented in class (5 minute talk) to get some extra points
Слайд 3Definition
Data analysis is the process of transforming raw data into
usable information, often presented in the form of a published
analytical article, in order to add value to the statistical output. (OECD)Data analysis is a process of inspecting, cleansing, transforming, and modeling data with the goal of discovering useful information, informing conclusions, and supporting decision-making (Wikipedia)
Both miss one important step – collecting data.
Most theories are about modeling, but 80% of the time a data scientist spends on data collection and cleansing
Слайд 4Data analysis techniques
Data mining
automatic discovery of useful information in
large data repositories
Descriptive statistics
summarizing features of data
Exploratory data analysis
finding
new features in dataConfirmatory data analysis
hypotheses testing
Predictive analytics
deriving predictions from data
Text analytics
extracting information from textual (i.e. unstructured) data
Слайд 5Two cultures of data analysis
Data is generated by a black
box
Input variables x (independent variables) go in one side (time
you spend on your home assignments)On the other side the response variables y come out (your grades)
Two main goals: prediction and information
Two approaches: data modeling culture and algorithmic modeling culture
Слайд 6Data modeling culture
Starts with assuming a data model for the
inside of the black box
The values of the parameters
are estimated from the data and the model then used for information and/or prediction Model validation: goodness-of-fit tests
Слайд 7Algorithmic modeling culture
Considers the inside of the box complex and
unknown
Tries to find a function f(x) - an algorithm that
operates on x to predict the responses yModel validation: predictive accuracy
Слайд 8Why do you need to learn data analysis
Valuable skill that
is highly remunerative
Things sometimes are not as obvious as they
seem at first sightAbility to verify results produced by your colleagues
The only way to make scientific contribution and verify theories, especially in social sciences
Слайд 9Data manipulation by Tim Cook
https://www.statschat.org.nz/2013/09/11/cumulative-totals-tend-to-increase/
Слайд 10Even academic superstars may be wrong
http://theconversation.com/the-reinhart-rogoff-error-or-how-not-to-excel-at-economics-13646
Слайд 11A lot of fraud in science (especially in social sciences)
https://www.financial-math.org/blog/2015/10/is-research-in-finance-and-economics-reproducible/
Слайд 12Random chance plays a huge role in social sciences
http://www.tylervigen.com/spurious-correlations
Слайд 15Intuition might be wrong, part 2
Monty Hall problem
https://en.wikipedia.org/wiki/Monty_Hall_problem
Humans vs birds: birds
win (Herbranson, 2010)
Слайд 16R
R is a language of statistical computing
Modern social sciences speak
mostly this language (and Python as well)
R download link: https://cran.r-project.org
RStudio
download: https://www.rstudio.com/products/rstudio/download/#download 
Слайд 17P.S.
Calling Bullshit is a highly recommended online course at the
University of Washington http://callingbullshit.org/syllabus.html#Introduction
