kaksivuoromalli

Kaksivuoromalli refers to a model designed to describe data that exhibit two distinct peaks or modes. It is commonly implemented as a mixture model, consisting of two underlying components that together generate the observed distribution. The term is used in statistics, signal processing, and related fields to capture bimodality in data.

The simplest instantiation is a two-component Gaussian mixture. The density is f(x) = w1 φ(x; μ1, σ1) + w2 φ(x; μ2, σ2), where φ is the normal density, and w1 and w2 are nonnegative weights with w1 + w2 = 1. More generally, the two components can be any parametric densities; the key feature is bimodality, with each component contributing a peak in different regions of the data space.

Parameters are typically estimated by maximum likelihood using the expectation-maximization (EM) algorithm, or by Bayesian methods with prior distributions. Common challenges include label switching (identifiability of the two components), selecting the correct number of components, and sensitivity to starting values or model misspecification.

Applications span statistics and data analysis, spectroscopy and chromatography for resolving overlapping peaks, image analysis for segmenting regions with two dominant intensities, and economics or psychology where data reflect two underlying states or regimes. The concept is related to broader themes of mixture models and bimodal distributions, and the term is widely used in Finnish scientific literature and software for modeling bimodal data.