arXiv subjects

Torbjörn Lundh

Publications and source records attributed to Torbjörn Lundh.

7 recordsLinked to original sources

arXiv2026

Evaluating infectious disease forecasts in a cost-loss situation

In order for epidemiological forecasts to be useful for decision-makers the forecasts need to be properly validated and evaluated. Although several metrics fore evaluation have been proposed and used none of them account for the potential costs and losses that the decision-maker faces. We have adapted a decision-theoretic framework to an epidemiological context which assigns a Value Score (VS) to each model by comparing the expected expense of the decision-maker when acting on the model forecast to the expected expense obtained from acting on historical event probabilities. The VS depends on the cost-loss ratio and a positive VS implies added value for the decision-maker whereas a negative VS means that historical event probabilities outperform the model forecasts. We apply this framework to a subset of model forecasts of influenza peak intensity from the FluSight Challenge and show that most models exhibit a positive VS for some range of cost-loss ratios. However, there is no clear relationship between the VS and the original ranking of the model forecasts obtained using a modified log score. This is in part explained by the fact that the VS is sensitive to over- vs. under-prediction, which is not the case for standard evaluation metrics. We believe that this type of context-sensitive evaluation will lead to improved utilisation of epidemiological forecasts by decision-makers.

q-bio.QM↗

arXiv2025

Evaluation of respiratory disease hospitalisation forecasts using synthetic outbreak data

Forecasts of hospitalisations of infectious diseases play an important role for allocating healthcare resources during epidemics and pandemics. Large-scale analysis of model forecasts during the COVID-19 pandemic has shown that the model rank distribution with respect to accuracy is heterogeneous and that ensemble forecasts have the highest average accuracy. Building on that work we generated a maximally diverse synthetic dataset of 324 different hospitalisation time-series that correspond to different disease characteristics and public health responses. We evaluated forecasts from 14 component models and 6 different ensembles. Our results show that component model accuracy was heterogeneous and varied depending on the current rate of disease transmission. Going from 7 day to 14 day forecasts mechanistic models improved in relative accuracy compared to statistical models. A novel adaptive ensemble method outperforms all other ensembles, but is closely followed by a median ensemble. We also investigated the relationship between ensemble error and variability of component forecasts and show that the coefficient of variation is predictive of future error. Lastly, we validated the results on data from the COVID-19 pandemic in Sweden. Our findings have the potential to improve epidemic forecasting, in particular the ability to assign confidence to ensemble forecasts at the time of prediction based on component forecast variability.

q-bio.PE↗

arXiv2021

Predicting regional COVID-19 hospital admissions in Sweden using mobility data

The transmission of COVID-19 is dependent on social contacts, the rate of which have varied during the pandemic due to mandated and voluntary social distancing. Changes in transmission dynamics eventually affect hospital admissions and we have used this connection in order to model and predict regional hospital admissions in Sweden during the COVID-19 pandemic. We use an SEIR-model for each region in Sweden in which the infectivity is assumed to depend on mobility data in terms of public transport utilisation and mobile phone usage. The results show that the model can capture the timing of the first and beginning of the second wave of the pandemic. Further, we show that for two major regions of Sweden models with public transport data outperform models using mobile phone usage. The model assumes a three week delay from disease transmission to hospitalisation which makes it possible to use current mobility data to predict future admissions.

q-bio.PE↗

arXiv2020

Mathematical models for pain: a systematic review

There is no single prevailing theory of pain that explains its origin, qualities, and alleviation. Although many studies have investigated various molecular targets for pain management, few have attempted to examine the etiology or working mechanisms of pain through mathematical or computational techniques. In this systematic review, we identified mathematical and computational approaches for characterizing pain. The databases queried were Science Direct and PubMed, yielding 560 articles published prior to January 1st, 2020. After screening for inclusion of mathematical or computational models of pain, 31 articles were deemed relevant. Most of the reviewed articles utilized classification algorithms to categorize pain and no-pain conditions. We found the literature heavily focused on the application of existing models or machine learning algorithms to identify the presence or absence of pain, rather than to explore features of pain that may be used for diagnostics and treatment. Although understudied, the development of mathematical models may augment the current understanding of pain by providing directions for testable hypotheses of its underlying mechanisms.

q-bio.NC↗

arXiv2016

Estimating the probability of coexistence in cross-feeding communities

The dynamics of many microbial ecosystems are driven by cross-feeding interactions, in which metabolites excreted by some species are metabolised further by others. The population dynamics of such ecosystems are governed by frequency-dependent selection, which allows for stable coexistence of two or more species. We have analysed a model of cross-feeding based on the replicator equation, with the aim of establishing criteria for coexistence in ecosystems containing three species, given the information of the three species' ability to coexist in their three separate pairs, i.e. the long term dynamics in the three two-species component systems. The triple-system is studied statistically and the probability of coexistence in the species triplet is computed for two models of species interactions. The interaction parameters are modelled either as stochastically independent or organised in a hierarchy where any derived metabolite carries less energy than previous nutrients in the metabolic chain. We differentiate between different modes of coexistence with respect to the pair-wise dynamics of the species, and find that the probability of coexistence is close to $\frac12$ for triplet systems with three pair-wise coexistent pairs and for so-called intransitive systems. Systems with two and one pair-wise coexistent pairs are more likely to exist for random interaction parameters, but are on the other hand much less likely to exhibit triplet coexistence. Hence we conclude that certain species triplets are, from a statistical point of view, rare, but if allowed to interact are likely to coexist. This knowledge might be helpful when constructing synthetic microbial communities for industrial purposes.

q-bio.PE↗

arXiv2004

War of attrition with implicit time cost

In the game-theoretic model war of attrition, players are subject to an explicit cost proportional to the duration of contests. We construct a model where the time cost is not explicitly given, but instead depends implicitly on the strategies of the whole population. We identify and analyse the underlying mechanisms responsible for the implicit time cost. Each player participates in a series of games, where those prepared to wait longer win with higher certainty but play less frequently. The model is characterised by the ratio of the winner's score to the loser's score, in a single game. The fitness of a player is determined by the accumulated score from the games played during a generation. We derive the stationary distribution of strategies under the replicator dynamics. When the score ratio is high, we find that the stationary distribution is unstable, with respect to both evolutionary and dynamical stability, and the dynamics converge to a limit cycle. When the ratio is low, the dynamics converge to the stationary distribution. For an intermediate interval of the ratio, the distribution is dynamically but not evolutionarily stable. Finally, the implications of our results for previous models based on the war of attrition are discussed.

q-bio.PE↗

arXiv2001

In search of an evolutionary coding style

In the near future, all the human genes will be identified. But understanding the functions coded in the genes is a much harder problem. For example, by using block entropy, one has that the DNA code is closer to a random code then written text, which in turn is less ordered then an ordinary computer code; see \cite{schmitt}. Instead of saying that the DNA is badly written, using our programming standards, we might say that it is written in a different style -- an evolutionary style. We will suggest a way to search for such a style in a quantified manner by using an artificial life program, and by giving a definition of general codes and a definition of style for such codes.

math.NA↗