Time series analysis and forecasting of cholera disease using discrete wavelet transform and seasonal autoregressive integrated moving average model

Entropy-metric estimation of the small data models with stochastic parameters

“…In both cases, these are stochastic linear programming problems with probabilistic equality constraints [ 34 , 35 ]. Having implemented in an analytical form the standard approach to solving such problems (see the authors' previous work on this topic [ 33 ]) we obtained optimal probability density functions classified as continuous differential functions, namely: where , , , , ; , , are the Lagrange multipliers, and corresponds to constraint [ 28 ], and , are correspond to constraints [ 27 , 28 ].…”

“…This can happen when the impact of individual instances on the quality metric is comparable to the combined impact of all other instances (this is typical for unbalanced classes or strong outliers). In cases where it is obvious that the IID hypothesis is limited, other approaches are used [ [24] , [25] , [26] , [27] , [28] , [29] ]: Markov chains, autoregressive models, stochastic decision trees, etc. However, there are borderline cases where it is not clear whether IID training can be applied or not.…”