Machine learning algorithms have become popular in diabetes research, especially within the scope of glucose prediction from continuous glucose monitoring (CGM) data. We investigated the design choices in case-based reasoning (CBR) approach to glucose prediction from the CGM data. Design choices were made with regards to the distance function (city-block, Euclidean, cosine, Pearson’s correlation), number of observations, and adaptation of the solution (average, weighted average, linear regression) used in the model, and were evaluated using five-fold cross-validation to establish the impact of each choice to the prediction error. Our best models showed mean absolute error of 13.35 ± 3.04 mg/dL for prediction horizon PH = 30 min, and 30.23 ± 6.50 mg/dL for PH = 60 min. The experiments were performed using the data of 20 subjects recorded in free-living conditions. The problem of using small datasets to test blood glucose prediction models and assess the prediction error of the model was also addressed in this paper. We proposed for the first time the methodology for estimation of the impact of the number of subjects (i.e., dataset size) on the distribution of the prediction error of the model. The proposed methodology is based on Monte Carlo cross-validation with the systematic reduction of subjects in the dataset. The implementation of the methodology was used to gauge the change in the prediction error when the number of subjects in the dataset increases, and as such allows the projection on the prediction error in case the dataset is extended with new subjects.
W artykule zaprezentowano metodę samo-uczenia mini-modeli (metodę MM) opartą na hiperbryłach w przestrzeni wielowymiarowej. Jest to metoda nowa i rozwojowa, będąca w trakcie intensywnych badań. Bazuje ona na próbkach pobieranych jedynie z lokalnego otoczenia punktu zapytania, a nie z obszarów odległych od tego punktu. Grupa punktów, używana w procesie uczenia mini-modelu jest ograniczona obszarem hiperbryły. na tak zdefiniowanym lokalnym otoczeniu punktu zapytania metoda MM w procesie uczenia oraz obliczania odpowiedzi można użyć dowolnej metody aproksymacji. W artykule przedstawiono algorytm uczenia i działania metody w przestrzeni wielowymiarowej bazujący na hipersferycznym układzie współrzędnych. Metodę przebadano na zbiorach danych wielowymiarowych, a wyniki porównano z innymi metodami bazującymi na próbkach.
EN
The article presents self-learning method of mini-models (MM-method) based on polytopes in multidimensional space. The method is new and is an object of intensive research. MM method is the instance based learning method and uses data samples only from the local neighborhood of the query point. Group of points which are used in the model-learning process is constrained by a polytope area. The MM-method can on a defined local area use any approximation algorithm to compute mini-model answer for the query point. The article describes a learning technique based on hyper-spherical coordinate system. The method was used in the modeling task with multidimensional datasets. The results of numerical experiments were compared with other instance based methods.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.