The increasing property for binomial and Poisson distributions

About 14 years old · traced to

Let MM be a binomially or Poisson-distributed random variable with GG-transform G(M)G(M), and let Φ\Phi denote the standard Gaussian distribution function. The increasing property. The function

m⟼P(M=m)Φ(G(m+1))−Φ(G(m))m\longmapsto\frac{P(M=m)}{\Phi\left(G(m+1)\right)-\Phi\left(G(m)\right)}

is increasing, while the function

m⟼P(M=m)Φ(G(m))−Φ(G(m−1))m\longmapsto\frac{P(M=m)}{\Phi\left(G(m)\right)-\Phi\left(G(m-1)\right)}

is decreasing. The source describes this as a stronger conjecture that implies the intersection property for binomial and Poisson distributions, but gives no resolution of the conjecture.

References

Primary source

Peter Harremoës and Gábor Tusnády, “Information Divergence is more chi squared distributed than the chi squared statistics”, arXiv:1202.1125 (2012).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.