Data Science/๊ฐœ๋…๊ณผ ์šฉ์–ด

There are two types of predictions in data science. Regression์€ numerical data๋ฅผ ์˜ˆ์ธกํ•˜๋Š”๋ฐ ์‚ฌ์šฉํ•˜๊ณ , Classification ์€ cateogorical data๋ฅผ ์˜ˆ์ธกํ•˜๋Š”๋ฐ ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค.  ์˜ˆ๋ฅผ ๋“ค์–ด, ์šฐ๋ฆฌ๊ฐ€ ์‚ฌ์šฉํ•˜๋Š” ์ด๋ฉ”์ผ์˜ ์ŠคํŒธ ๋ฉ”์ผํ•จ์ด ์žˆ์Šต๋‹ˆ๋‹ค.๋ฉ”์ผ์˜ ํ…์ŠคํŠธ ๋ฐ์ดํ„ฐ๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ์ŠคํŒธ์ธ์ง€ ์•„๋‹Œ์ง€, Yes or No ์— ํ•ด๋‹นํ•˜๋Š” Cateogorical variable์„ ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค. Input = Text / Output = Yes or No (Spam, Not Spam) Classification์— ๋Œ€ํ•ด์„œ ๋” ์ž์„ธํžˆ ์•Œ์•„๋ณด๊ธฐ ์ด์ „, ๊ฐ„๋‹จํ•˜๊ฒŒ Machine Learning์— ๋Œ€ํ•ด์„œ ์•Œ์•„๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.  Machine Learning Algorit..
In a simple linear regression line (LMS), the regression line can be expressed as following equation:y = ax + bwherey = The variable that you want to predict (์˜ˆ์ธกํ•˜๊ณ  ์‹ถ์€ ๊ฐ’) | Dependent variable (์ข…์† ๋ณ€์ˆ˜)x = The variable that you are using to predict (์˜ˆ์ธก์— ์‚ฌ์šฉํ•˜๋Š” ๊ฐ’) | Independent variable (๋…๋ฆฝ ๋ณ€์ˆ˜)a = Slope (๊ธฐ์šธ๊ธฐ)b = y-intercept (y ์ ˆํŽธ) ๊ทธ๋ ‡๋‹ค๋ฉด, y = ax+b ์—์„œ slope(a)์™€ y-intercept(b)๋Š” ์–ด๋–ป๊ฒŒ ๊ตฌํ•˜๋Š”์ง€ ์•Œ์•„๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค. Recall, r ..
When there are two numerical variables, there are  TrendPositive associationNegative association PatternAny discernible "shape" in the scatterLinear Non-linear Visualize, then quantify  The Correlation Coefficient rMeasures linear association. It is based on the standard units. r is defined as:The average of product of (x in standard units) and (y in standard units) ํ‘œ์ค€ ๋‹จ์œ„ x์™€ ํ‘œ์ค€ ๋‹จ์œ„ y์˜ ๊ณฑ์˜ ํ‰๊ท   In P..
Confidence Interval is the interval of estimates of a parameter.It's based on random sampling. ๊ฐ€์žฅ ๋„๋ฆฌ ์‚ฌ์šฉ๋˜๋Š” ๊ฒƒ์€ 95% Confidence Interval ์ž…๋‹ˆ๋‹ค. Here, '95%' is called the confidence level.it could be any percent between 0 - 100.Higher confidence level means wider intervals.  Confidence interval can be considered 'Good' if it contains the parameter.The confidence is in the process that creates the inter..
Chan Lee
'Data Science/๊ฐœ๋…๊ณผ ์šฉ์–ด' ์นดํ…Œ๊ณ ๋ฆฌ์˜ ๊ธ€ ๋ชฉ๋ก