쉐도잉 연습: Introduction to statistical prediction models - 영상으로 영어 말하기 배우기

연습 모드 선택

각 모드는 전체 화면 연습실로 열립니다.

125 문장
문장이 너무 짧거나 길면 Edit를 눌러 조정하세요.
1
In this video I will give you a very short introduction to what statistical prediction is and what statistical prediction models are.
2
And my goal of this video is to try to you to understand what a large language model,
3
which is the technology underlying these AI chatbots,
4
what they do and why they sometimes hallucinate or give you incorrect answers.
5
So let's start with a more effective example.
6
This is Claude III, the smallest model,
7
and this is something that I encountered when I was preparing material for a doctoral level entrepreneurship course.
8
And I asked what is the contribution of Israel Kirchner in modern entrepreneurship, and then I asked Claude to explain or give me his biography.
9
And it starts well, most of this is correct.
10
So Israel Kirchner was born in 1930s, and then the biography ends that he passed away at 2021 at the age of 90.
11
Well, he is very much alive still.
12
So if you google his name, you can view that the dead guy is giving a talk on YouTube in 2023.
13
So now the question is, why does the large language model,
14
Claude in this case, give us this incorrect answer that the person who is alive was actually died a couple years ago.
15
To understand why we get incorrect answers from these tools, we need to understand a bit about what they do.
16
So let's talk about statistical prediction a bit.
17
So here's our some data, then we're going to fit a statistical prediction model to this data.
18
So we have sales, a fictional company how many units they sell and this is how much they advertise.
19
This might be something expensive equipment and this is your advertisement spending thousands, let's say that it's over a year.
20
And we can see that if the company doesn't advertise anything they sell about 150,
21
130 something that units and then there is this declining returns to advertisement.
22
So a little bit of advertisement gets you from 150 to 200, but then going beyond 200 is much more difficult.
23
So we have diminishing returns.
24
The simplest way to model this data statistically is to use a regression model.
25
And regression model happens to be the simplest statistical prediction model.
26
The idea of a regression model is that we try to explain this data with a line.
27
And we draw the line through the data.
28
How the line is defined is not important for this talk, but it's a line that describes the data.
29
And this line is defined by two parameters.
30
So these parameters define a model.
31
So we are saying that this beta 0 gives us what is the the value of the dependent variable,
32
the predicted variable, the sales when ad spending is zero,
33
and then beta 1 gives us how many units the sales increases for each thousands of dollars spend on advertisement.
34
In machine learning, you might also see that these are called weights and biases.
35
So the beta one is weight because it tells how much ad spending affects sales,
36
and then beta zero is called the bias because it tells what is the overall level of sales, even if we don't do any advertisement.
37
And how the recursion line here is calculated is that we take the bias 158,
38
we add weight times add spending, and then if we want to generate observations from this model, we add some error.
39
So the idea of error is that there's some variation in the actual values around the predicted line.
40
So this equation without error here gives you the recursion line, and if you want to simulate observations from this data then you would add error.
41
So we can use this line to simulate more data.
42
So if we have, if we know ad spending, then we can just plot more predicted sales values here.
43
This works fairly well as long as you predict within samples.
44
So if you predict something that is, you have data, so our predictions work fairly well between zero and fifty thousand.
45
We are predicting a bit more here than what we are observing.
46
Problems occur when we start to extrapolate.
47
So if we are asking this model to predict how much we would sell if we actually increased our sales spending,
48
our advertisement spending, to a million.
49
Here's the data.
50
So we can predict what the sales would look like if we advertised for a million dollars.
51
Now this model is not very trustworthy for this kind of predictions.
52
First of all, it doesn't consider that advertisement has diminishing returns.
53
So the first dollar is more valuable than the tenth dollar and so on.
54
Also there are probably other constraints that come into play.
55
So So just increasing your advertisement 100-fold will not increase your sales 100-fold.
56
There are other constraints.
57
So the real value might be somewhere closer to 200 or 500 units than 1,500 units.
58
So extrapolation generally is not good in statistical models.
59
At least this line of gross extrapolation.
60
So we need to understand now that we calibrate the model using data, in this case a regression model.
61
It gives us some parameters and that parameters, they give us an equation and we can use the equation to calculate predictions.
62
The predictions are closed with original data as long as the values that we predict with our closed-grid original data.
63
All right, now let's take a look at how this same idea works with text.
64
So this is our training data and we are going to specify a small language model
65
and we try to predict more text like this.
66
And the idea of modeling language is that instead of looking at what is the predicted sales are given advertisements.
67
We look at what is the predicted word given the previous word.
68
And this kind of model will be called biogram, but it's not super important to understand how it works.
69
But the general idea is that when we start to train this kind of a prediction model that works with language, we take a look at, for example, the word the.
70
And the is here, and the is followed by cat three times.
71
It's followed by dog once and mat once.
72
So we can say that when we see the word the, then 60% of the case we have cat as the next word, and 20% of the case we have dog,
73
20% case we have mat.
74
So that gives us kind of like these probabilities and we can start simulating new data from this model.
75
Then after cat we have sat, the and ran.
76
So all of these have 33% probability and so on.
77
So we can calculate the probabilities.
78
This is the word here on this on rows and then columns is the next word.
79
We can start using this model to generate data.
80
And this is what large language models do.
81
So we have an initial, let's call it prompt.
82
So this is what we would write to chat GPT, for example.
83
We just start with the.
84
And then what the model looks at, it looks that, okay, the word is the.
85
When we predict the next word, we have cat, 60% dog, and then 20% mat.
86
And it randomly picks one of these.
87
This has 60% probability 20% and 20% and it picks cat by random.
88
Then after cat we have either ran, sat and the with equal probabilities,
89
the model just picks randomly sat.
90
And after sat we always have on, so that is what the model predict,
91
and so on until it predicts away, which is the last word.
92
and this is how language models work.
93
So we have a word and then we predict what is the next word.
94
Then we take the predicted word and we predict the following word.
95
This is a pretty simple thing and this story doesn't really make sense.
96
The cat sat on the cat, sat on the mat, the cat sat on, the cat ran away.
97
But it looks somewhat similar to the original data.
98
We can make this a lot better if we start to look at broader context.
99
So instead of looking at just previous word, we could take a look at the two previous words or three previous words.
100
We can increase the context length.
101
So this small language model has context length of one word, and then we have 81 parameters because we have 81 probabilities.
102
I just didn't include the zeros in this matrix.
103
This is how ChatGPT works.
104
So it looks what I have written this far, what is the likely next word.
105
ChatGPT, Claude, Gemini, they all work with the same principle, they are just a lot bigger.
106
So instead of 81 parameters like we have here, these models are in the hundreds of billions,
107
probably the trillions of parameters and the context length is not one word, but it is something called a token,
108
which is like a word part or a letter combination.
109
And for example, Claude has 100,000 tokens.
110
That's like multiple books that it looks at when it tries to predict the next word.
111
Now we get to look at why did Claude predict incorrectly.
112
And this is because of XAAAs.
113
So the training data for Claude
114
when it was trained using like most of the internet does not contain information about the death of Kirchner
115
and it's natural because he is still alive.
116
And here we have that the person died, was born in 1930,
117
and the training data also contains lots of biographies of people born in 1930
118
and most of those biographies end with explaining when the person passed.
119
So the language model does not have information about the passing of Kirchner,
120
but it has seen a general pattern that when the biography starts by telling that the person was born in 1930,
121
then the likely ending for the biography is just that the person died somewhere in 2020s for example.
122
So this is the basic idea of statistical prediction.
123
You calibrate the data, calibrate a model based on data, and then you try to predict new observations.
124
In large language models you predict the next word or the next word part.
125
In recursive models you predict a value, for example.

앱 다운로드

유창하게 말하는 데 필요한 모든 것

AI 발음 평가문장마다 점수 확인
IPA 발음 연습모든 소리 마스터
단어 학습나만의 단어장 만들기
단어 게임놀면서 외우기

같은 채널

FOR YOU

추천 영상

이 레슨의 어휘와 말하기 포인트

이 C1 수준 말하기 레슨은 영상 “Introduction to statistical prediction models”을(를) 바탕으로 합니다. 가장 자주 반복되는 단어는 다음과 같습니다: model, word, predict, prediction, cat. 이 영상에는 섀도잉할 문장 125개와 단어 1690개가 있습니다. 말하는 구간의 길이는 11:51입니다. 화자는 분당 약 143단어의 일정한 속도로 말해서 섀도잉하기에 편한 속도입니다. 단어의 86%가 영어에서 가장 많이 쓰이는 3,000단어에 속합니다. 나머지는 연습 전에 미리 확인해 두세요.

이 영상의 핵심 어휘

영상에서 가장 어려운 단어 15개를 발음, 뜻과 함께 정리했습니다.

단어발음뜻
predict 동사/pɹɪˈdɪkt/예언하다, 예측하다
prediction 명사/pɹɪˈdɪkʃən/예언, 예측
advertisement 명사/ˈædvɚˌtaɪzmənt/광고
parameter 명사/pəˈɹæm.ə.tɚ/파라미터, 매개변수
biography 명사/baɪˈɑɡɹəfi/전기
probability 명사/ˌpɹɑ.bəˈbɪl.ə.ti/확률
regression 명사/ɹiːˈɡɹɛʃ.ən/회귀
calculate 동사/ˈkælkjʊleɪt/정산하다, 계산하다
incorrect 형용사/ˌɪn.kəˈɹɛkt/그른, 잘못된
equation 명사/ɪˈkweɪ.ʒən/방정식
token 명사/ˈtoʊkən/토큰
fold 동사/ˈfəʊld/개다, 접다
chatbot 명사/ˈtʃætbɑt/챗봇
prompt 동사/pɹɑmpt/부추기다
tenth 형용사/tɛnθ/열째

이 영상의 문법

화자가 가장 많이 쓰는 문형을 영상 속 실제 표현과 함께 정리했습니다.

문형영상 속 표현
수동태 be + 과거분사 — 누가 하는지보다 무슨 일이 일어나는지에 초점was born · was actually died · is defined
현재완료 have/has + 과거분사 — 과거의 일이 지금도 관련이 있을 때have sat · have written · has seen
관계절 who / which + 절 — 사람이나 사물에 대한 추가 정보person who is · away, which is

주의할 발음

화자는 doesn't, didn't, don't 같은 축약형과 약화된 형태를 6번 사용합니다. 들리는 대로 짧게 발음하세요.

  • “th” 소리: trustworthy /ˈtɹʌst.wɜː.ði/, tenth /tɛnθ/
  • “sh”와 “zh” 소리: prediction /pɹɪˈdɪkʃən/, regression /ɹiːˈɡɹɛʃ.ən/, equation /ɪˈkweɪ.ʒən/, diminish /dɪˈmɪnɪʃ/, extrapolation /ɛkˌstɹæp.əˈleɪ.ʃən/
  • 긴 단어 — 강세 위치에 주의: advertisement /ˈædvɚˌtaɪzmənt/, parameter /pəˈɹæm.ə.tɚ/, biography /baɪˈɑɡɹəfi/, probability /ˌpɹɑ.bəˈbɪl.ə.ti/, extrapolation /ɛkˌstɹæp.əˈleɪ.ʃən/

이 영상으로 연습하는 방법

  1. 먼저 말하지 않고 영상을 끝까지 듣고 모르는 단어를 적어 둡니다.
  2. 보통 속도로 한 문장씩 섀도잉하고, 화자의 리듬과 맞을 때까지 반복합니다.
  3. 자신의 목소리를 녹음해 원본과 비교하고, predict, prediction, advertisement 같은 단어에 특히 주의합니다.

쉐도잉이란? 영어 실력을 빠르게 키우는 과학적 방법

쉐도잉(Shadowing)은 원래 전문 통역사 훈련을 위해 개발된 언어 학습 기법으로, 다언어 학자인 Dr. Alexander Arguelles에 의해 대중화된 방법입니다. 핵심 원리는 간단하지만 매우 강력합니다: 원어민의 영어를 들으면서 1~2초의 짧은 지연으로 즉시 소리 내어 따라 말하는 것——마치 '그림자(shadow)'처럼 화자를 따라가는 것입니다. 문법 공부나 수동적인 청취와 달리, 쉐도잉은 뇌와 입 근육이 동시에 실시간으로 영어를 처리하고 재현하도록 훈련합니다. 연구에 따르면 이 방법은 발음 정확도, 억양, 리듬, 연음, 청취력, 말하기 유창성을 크게 향상시킵니다. IELTS 스피킹 준비와 자연스러운 영어 소통을 원하는 분들에게 특히 효과적입니다.

섀도잉 방법: 단계별 전체 가이드 읽기 →