search

tensorflow / probability

Probabilistic reasoning and statistical analysis in TensorFlow
https://www.tensorflow.org/probability/
Apache License 2.0
4.27k stars 1.1k forks source link

Kappa-Generalized distribution #1607

Open leandrolcampos opened 2 years ago

leandrolcampos commented 2 years ago

Hello TFP community,

I’d like to work on the implementation of the Kappa-Generalized distribution [1-3]. It's based on the one-parameter generalization of the exponential function proposed in [4] and can be used, for instance, to describe the distribution and dispersion of income within a population. As noted by [3], this distribution is able to "describe the whole spectrum of incomes, from the low-middle income region up to the high-income Pareto power-law regime". It can also be applied for epidemiological analysis [5].

Can I work on this task? And any suggestion about how to name this distribution in TFP?

All the best,

[1] Clementi, Fabio, Mauro Gallegati, and G. Kaniadakis. "κ-generalized statistics in personal income distribution." The European Physical Journal B 57.2 (2007): 187-193. Link

[2] Clementi, Fabio, et al. "The κ-generalized distribution: A new descriptive model for the size distribution of incomes." Physica A: Statistical Mechanics and its Applications 387.13 (2008): 3201-3208. Link

[3] Clementi, Fabio, Mauro Gallegati, and G. Kaniadakis. "A κ-generalized statistical mechanics approach to income analysis." Journal of Statistical Mechanics: Theory and Experiment 2009.02 (2009): P02037. Link

[4] Kaniadakis, G. "Non-linear kinetics underlying generalized statistics." Physica A: Statistical mechanics and its applications 296.3-4 (2001): 405-425. Link

[5] Kaniadakis, Giorgio, et al. "The κ-statistics approach to epidemiology." Scientific Reports 10.1 (2020): 1-14. Link

leandrolcampos commented 2 years ago

This distribution is also known as Kaniadakis Weibull distribution (https://en.wikipedia.org/wiki/Kaniadakis_Weibull_distribution).

brianwa84 commented 2 years ago

Looks like a nice addition, KaniadakisWeibull is probably the name we'd prefer, seems most specific.

On Thu, Aug 25, 2022, 9:22 AM Leandro Campos @.***> wrote:

This distribution is also known as Kaniadakis Weibull distribution ( https://en.wikipedia.org/wiki/Kaniadakis_Weibull_distribution).

— Reply to this email directly, view it on GitHub https://github.com/tensorflow/probability/issues/1607#issuecomment-1227252693, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFJFSI7GQEZINRFOT36IVZ3V25XRFANCNFSM57S4ZHCQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

leandrolcampos commented 2 years ago

Thanks for the answer, @brianwa84