Thuật ngữ AI
Từ điển đầy đủ về Trí tuệ nhân tạo
Frequency distribution
Tabular or graphical representation of the number of occurrences of each value or class of a quantitative or qualitative variable, allowing visualization of the data structure.
Histogram
Bar chart with contiguous bars representing the distribution of a continuous quantitative variable, where the area of each bar is proportional to the frequency of observations in the corresponding interval.
Box plot
Statistical diagram summarizing the distribution of a variable through five indicators: minimum, first quartile, median, third quartile, and maximum, revealing symmetry and the presence of extreme values.
Quartiles
Values that divide an ordered dataset into four equal parts, where Q1 represents the 25th percentile, Q2 the median (50th percentile), and Q3 the 75th percentile.
Standard deviation
Measure of dispersion calculated as the square root of the variance, quantifying the average dispersion of observations around the mean of the distribution.
Variance
Average of the squared deviations from the mean, measuring the dispersion of data and forming the basis for calculating the standard deviation in univariate analysis.
Arithmetic mean
Central tendency indicator obtained by dividing the sum of all observations by their total number, sensitive to extreme values and appropriate for symmetric data.
Median
Central value separating the ordered distribution into two equal parts, robust to extreme values and particularly informative for asymmetric distributions.
Mode
Value or category appearing with the highest frequency in a distribution, which can be unique, multiple, or non-existent depending on the data structure.
Range
Elementary measure of dispersion calculated as the difference between the maximum and minimum value of a variable, providing an initial assessment of data variability.
Coefficient of variation
Ratio of the standard deviation to the mean, expressed as a percentage, allowing comparison of relative dispersion between variables of different scales in univariate analysis.
Skewness
Asymmetry coefficient measuring the degree of inclination of a distribution relative to its symmetry, distinguishing left-skewed (negative) from right-skewed (positive) distributions.
Bar chart
Graphical representation of qualitative or discrete quantitative variables where the height of each bar is proportional to the frequency or count of each category.
Pie chart
Graph divided into angular sectors proportional to the relative frequencies of a qualitative variable, visualizing the composition of a whole in percentages.
Density plot
Smoothed curve estimating the probability density function of a continuous variable, providing an alternative visualization to the histogram for identifying modes and the shape of the distribution.
Interquartile range
Robust measure of dispersion calculated as the difference between the third and first quartile (Q3-Q1), representing the range containing 50% of the central observations.
Pearson's skewness coefficient
Skewness index based on the difference between mean and median, normalized by the standard deviation, providing an alternative measure of skewness that is particularly intuitive.
Outliers
Extreme observations that deviate significantly from the general pattern of the data, identified by methods such as the interquartile range (1.5×IQR) or Z-score.