rassta: Raster-based Spatial Stratification Algorithms

Spatial stratification of landscapes allows for the development of efficient sampling surveys,the inclusion of domain knowledge in data-driven modeling frameworks, and the production of information relating the spatial variability of response phenomena to that of landscape processes. This work presents the rassta package as a collection of algorithms dedicated to the spatial stratification of landscapes, the calculation of landscape correspondence metrics across geographic space, and the application of these metrics for spatial sampling and modeling of environmental phenomena. The theoretical background of rassta is presented through references to several studies which have benefited from landscape stratification routines. The functionality of rassta is presented through code examples which are complemented with the geographic visualization of their outputs.

Observing Brief and Irregular Behaviour of Animals⁠ —The Validity of Short Observation Periods

There are efficient sampling methods to accurately estimate behaviour with a moderate or long duration. For short behaviour, observing animals continuously is recommended although there is no recommended minimum observation time. In most studies, sampling method and observation time per day is determined by practical considerations. Thus, this study analysed the validity of behavioural observations in different observation periods using continuous sampling (CS) or time sampling (TS) based on biting behaviour. Tail-biting and ear-biting of weaned piglets in six pens were continuously observed for 12 h per day for 4 days to form a reference. Shorter observation periods of CS and TS were simulated by taking subsets of this reference. The amount of behaviour per hour of each observation period was compared to the reference and to other observation period of the same kind and length. Four different measurements were defined to calculate accuracy scores (AS; 0–1; higher values are better). Comparison to the reference shows better AS for observation periods with longer observation time in total (0.5 h of CS: 0.2; 6 h of CS: 0.6). Additionally, TS covers longer time periods without decreasing AS. However, focus on activity time results in an overestimation of irregular behaviour. Comparing AS among observation periods of the same kind and length show overall low agreement. This study indicated problems of different observation periods of CS and TS to accurately estimate behaviour. Therefore, validity of behavioural observations should be analysed in greater detail to determine optimal sampling methods.

Deep learning-based detection of tax frauds: an application to property acquisition tax

PurposeSampling taxpayers for audits has always been a major concern for policymakers of tax administration. The purpose of this study is to propose a systematic method to select a small number of taxpayers with a high probability of tax fraud.Design/methodology/approachAn efficient sampling method for taxpayers for an audit is investigated in the context of a property acquisition tax. An autoencoder, a popular unsupervised learning algorithm, is applied to 2,228 tax returns, and reconstruction errors are calculated to determine the probability of tax deficiencies for each return. The reasonableness of the estimated reconstruction errors is verified using the Apriori algorithm, a well-known marketing tool for identifying patterns in purchased item sets.FindingsThe sorted reconstruction scores are reasonably consistent with actual fraudulent/non-fraudulent cases, indicating that the reconstruction errors can be utilized to select suspected taxpayers for an audit in a cost-effective manner.Originality/valueThe proposed deep learning-based approach is expected to be applied in a real-world tax administration, promoting voluntary compliance of taxpayers, and reinforcing the self-assessing acquisition tax system.