Foundation of data science: Computational thinking with python

1. Section 1: Introduction

What is the data science? 

- drawing useful conclusions from data using computation as our primary tool.

And data science, as a practice, has three core activities: 

1. Exploration is figuring out what patterns exist in the data.
2. When you have many observations about some phenomenon, what can you conclude about the phenomenon itself?
3. Oftentimes instead of just looking at large tables of numbers, we'll draw data visualizations because it's much easier to interpret lot of information at once if it's portrayed in some kind of visual way.
4. Once we've found a pattern, we need to perform statistical inference, and that's because some patterns are there just by chance and some are there because they're a reflection of some underlying process that's really interesting about the world.
5. So the goal of statistical
6. inference is to quantify whether
7. the patterns that we observe
8. during the exploration phase are
9. reliable.
10. If we collected more data, would
11. we see this pattern again or
12. not?
13. The primary tool we have is
14. randomization because by
15. simulating random processes, we
16. can see what kinds of patterns
17. appear just by chance.
18. And if the pattern we observe is
19. not the kind of thing that could
20. just appear by chance, then we
21. can conclude that it's because
22. of some robust or reliable
23. pattern in the underlying
24. phenomenon we want to study.
25. And finally, we'll perform
26. prediction.
27. This is where we have partial
28. information about something we
29. want to know, and we want to
30. guess about the things we don't
31. know yet.
32. Here we're making informed
33. guesses, quantitative guesses
34. using a discipline called
35. machine learning.
36. Normally when we write programs,
37. we just focus on the particular
38. logic of what the computer
39. should do, but machine learning
40. is about not programming every
41. detail, but instead using the
42. data to make decisions or choice
43. within that program.
44. So when we write a program, for
45. instance, to recognize speech or
46. automatically translate
47. languages or control a car or a
48. robot, we don't actually write
49. down all the details of what to
50. do, but instead use examples
51. from the world to help computers
52. automatically learn how to
53. behave.
54. And that's a form of prediction,
55. one that we'll talk about in
56. this course.
57. And these three stages
58. correspond to how we'll approach
59. the material in this course.
60. We'll first talk about how to
61. identify patterns, then we'll
62. talk about quantifying whether
63. those patterns are reliable.
64. And finally, based on the
65. patterns we've discovered, the
66. reliable ones can help us make
67. informed guesses about the
68. information that we wish we
69. knew.
70. Once you can do all that,
71. you're well on your way to being
72. a data scientist.
73. Now in the process of doing all
74. these things, it's important
75. that you learn how to program a
76. computer, because computing
77. underlies each step of the way
78. and learning to program is just
79. an essential part of
80. participating in this
81. discipline.

2. Section 2: