This document provides a comprehensive overview of essential statistical concepts and their applications in data science. It covers a wide range of topics, from basic descriptive statistics to more advanced metrics like skewness, kurtosis, and correlation, complete with instructional visualizations.
Descriptive Statistics
Measures of Center: Mean, median, and mode are used to describe the central tendency of a dataset.
Normal Distribution: A bell-shaped probability distribution that is widely used in statistics.
Measures of Spread: Variance, standard deviation, and quartiles are used to describe the spread or dispersion of data points around the mean.
Measures of Shape: Skewness and kurtosis describe the overall topography of a distribution:
Skewness:
A measure of the asymmetry of a probability distribution.
Indicates the direction and degree to which a distribution leans towards one side of the mean.
A symmetrical distribution has a skewness of zero.
A distribution skewed to the right (positive skew) has a longer tail on the right side.
A distribution skewed to the left (negative skew) has a longer tail on the left side.
Kurtosis:
A measure of the "tailedness" of a probability distribution.
Describes how concentrated the data points are around the mean compared to the tails.
A mesokurtic distribution has a kurtosis similar to a normal distribution.
A leptokurtic distribution has a higher peak and fatter tails, indicating more extreme values.
A platykurtic distribution has a lower peak and thinner tails, indicating fewer extreme values.
Correlation
Covariance: A measure of how two variables change together.
Pearson's r: A standardized measure of the linear relationship between two continuous variables, ranging from -1 to +1.
Other Important Concepts
Inference: The process of drawing conclusions about a population based on a sample of data.
Variable: Any characteristic or attribute that can be measured or counted.
Outliers: Data points that are significantly different from other observations in the dataset.
Moments: Statistical measures that describe the shape of a distribution, including mean, variance, skewness, and kurtosis.
Train-Test Split: A technique used to split a dataset into training and testing sets for machine learning model evaluation.