Bernstein inequalities

E354909

Bernstein inequalities are fundamental results in approximation theory and probability that provide bounds on the derivatives or deviations of functions and random variables under certain smoothness or moment conditions.

All labels observed (7)

How this entity was disambiguated

Statements (49)

Predicate Object
instanceOf approximation theory result ⓘ
mathematical inequality ⓘ
probability inequality ⓘ
appliesTo functions with bounded derivatives ⓘ
independent random variables ⓘ
polynomials on a bounded interval ⓘ
characterizedBy dependence on variance and uniform bound of random variables ⓘ
exponential decay of tail probabilities ⓘ
trade-off between variance term and maximal bound term ⓘ
field approximation theory ⓘ
probability theory ⓘ
givesUpperBoundOn supremum norm of derivatives of polynomials ⓘ
tail probabilities of sums of random variables ⓘ
hasVariant Bernstein inequality for polynomials ⓘ
Bernstein inequality on compact sets ⓘ
Bernstein inequality on the circle ⓘ
Bernstein inequality on the real line ⓘ
Bernstein-type concentration inequality ⓘ
classical Bernstein inequality for sums of independent random variables ⓘ
matrix Bernstein inequality ⓘ
namedAfter Sergei Natanovich Bernstein ⓘ
provides bounds on derivatives of polynomials ⓘ
bounds on deviations of sums of random variables ⓘ
relatedTo Azuma–Hoeffding inequality ⓘ
Bennett inequality ⓘ
Bernstein polynomials ⓘ
Chernoff bound ⓘ
Hoeffding inequality ⓘ
linked to: Chernoff bound

Markov brothers' inequalities ⓘ
requiresCondition bounded random variables or bounded moments ⓘ
finite variance ⓘ
smoothness conditions on functions ⓘ
typicalAssumption bounded support or sub-exponential tails ⓘ
independence of summands ⓘ
zero-mean random variables ⓘ
usedFor concentration of measure ⓘ
confidence bounds for random sums ⓘ
convergence rates of estimators ⓘ
empirical process theory ⓘ
error bounds in approximation theory ⓘ
high-dimensional statistics ⓘ
large deviations estimates ⓘ
non-asymptotic probability bounds ⓘ
statistical learning theory ⓘ
uniform approximation of functions ⓘ
usedIn analysis of randomized algorithms ⓘ
machine learning generalization bounds ⓘ
risk bounds for empirical risk minimization ⓘ
signal processing and harmonic analysis ⓘ

How these facts were elicited

Referenced by (8)

Full triples — surface form annotated when it differs from this entity's canonical label.

Felix Bernstein → notableWork → Bernstein inequalities ⓘ
Felix Bernstein → notableConcept → Bernstein inequalities in approximation theory ⓘ
linked to: Bernstein inequalities
Bernstein inequalities → hasVariant → Bernstein inequality for polynomials ⓘ
linked to: Bernstein inequalities
Bernstein inequalities → hasVariant → Bernstein-type concentration inequality ⓘ
linked to: Bernstein inequalities
Bernstein inequalities → hasVariant → Bernstein inequality on the real line ⓘ
linked to: Bernstein inequalities
Bernstein inequalities → hasVariant → Bernstein inequality on the circle ⓘ
linked to: Bernstein inequalities
Bernstein inequalities → hasVariant → Bernstein inequality on compact sets ⓘ
linked to: Bernstein inequalities
Sergei Natanovich Bernstein → notableConcept → Bernstein inequalities ⓘ