7.11
Applications de l’intégration aux fonctions de densité de probabilité
Les distributions de probabilité continues sont utilisées pour modéliser les variables aléatoires pouvant prendre n’importe quelle valeur réelle dans…
Les distributions de probabilité continues modélisent des variables aléatoires pouvant prendre n’importe quelle valeur réelle dans une plage. Par exemple, la taille des femelles adultes peut être de 163,5, 165,25 centimètres, ou n’importe quelle valeur intermédiaire. Cela fait de la hauteur une variable aléatoire continue.
La probabilité de cette variable est décrite à l’aide d’une fonction densité de probabilité — une courbe lisse sur l’étendue de la variable. L’unité de cette densité est l’inverse de l’unité de la variable.
La probabilité totale que la variable se situe dans un intervalle spécifique est déterminée en calculant la surface sous la courbe pour cette plage. Par exemple, la probabilité que la taille d’une femme soit comprise entre 150 et 170 centimètres est déterminée en intégrant la fonction de densité sur cette plage.
Si le résultat est de 0,75, cela signifie que 75 % des femmes de la population ont une taille dans cette période.
Mais pour une hauteur exacte, la probabilité est nulle car l’intégration sur un seul point donne une surface nulle.
Une fonction de densité de probabilité valide est toujours non négative et s’intègre à 1 sur toute la plage des valeurs possibles pour la variable aléatoire.
View the full transcript and gain access to JoVE Core videos
Q1: What is a continuous random variable and how does it differ from discrete values?
A continuous random variable can take any real value within a specified range, rather than isolated countable values. For example, a woman's height might be 163.5, 165.25, or any value in between centimeters. This contrasts with discrete variables that have specific, separated outcomes. Continuous variables exist on a continuum and require integration to calculate probabilities over intervals.
Q2: How does integration relate to finding probabilities in continuous distributions?
Integration calculates the area under a probability density function curve over a specified interval, which represents the probability that a variable falls within that range. For instance, integrating the height density function between 150 and 170 centimeters yields the probability that a randomly selected woman has height in that interval. The area under the curve directly translates to probability.
Q3: Why is the probability of a continuous variable taking an exact single value always zero?
The probability of an exact value is zero because integration over a single point produces zero area under the curve. In continuous distributions, probability is defined only for intervals, not individual points. This fundamental property distinguishes continuous probability from discrete probability, where specific outcomes can have non-zero probabilities.
Q4: What properties must a valid probability density function satisfy?
A valid probability density function must be non-negative for all values in its domain and must integrate to 1 across the entire range of possible values. These properties ensure the function accurately represents a probability distribution and that total probability across all outcomes equals 1, maintaining mathematical consistency.
Q5: How do you interpret the result when integrating a probability density function over an interval?
The integral result represents the probability as a decimal or percentage. If integrating a height density function from 150 to 170 centimeters yields 0.75, this means 75% of the population has heights within that interval. The numerical result directly translates to the likelihood of the variable falling in that range.
Q6: What is the unit of measurement for a probability density function?
The unit of a probability density function is the reciprocal of the variable's unit. For height measured in centimeters, the density function has units of inverse centimeters. This reciprocal relationship ensures that when the density is multiplied by an interval width during integration, the result is a dimensionless probability.
Q7: How does the area under a probability density curve relate to real-world probability outcomes?
The area under the probability density curve over an interval directly represents the proportion of the population or outcomes within that range. For example, if the area under a height distribution curve between two values is 0.75, then 75% of individuals in the population have heights in that interval. Area and probability are equivalent in continuous distributions.