Deep self-supervised transformation learning for leukocyte classification

The scarcity of training annotation is one of the major challenges for the application of deep learning technology in medical image analysis. Recently, self-supervised learning provides a powerful solution to alleviate this challenge by extracting useful features from a large number of unlabeled training data. In this article, we propose a simple and effective self-supervised learning method for leukocyte classification by identifying the different transformations of leukocyte images, without requiring a large batch of negative sampling or specialized architectures. Specifically, a convolutional neural network backbone takes different transformations of leukocyte image as input for feature extraction. Then, a pretext task of self-supervised transformation recognition on the extracted feature is conducted by a classifier, which helps the backbone learn useful representations that generalize well across different leukocyte types and datasets. In the experiment, we systematically study the effect of different transformation compositions on useful leukocyte feature extraction. Compared with five typical baselines of self-supervised image classification, experimental results demonstrate that our method performs better in different evaluation protocols including linear evaluation, domain transfer, and finetuning, which proves the effectiveness of the proposed method.     

Adaptive machine learning for protein engineering

Machine-learning models that learn from data to predict how protein sequence encodes function are emerging as a useful protein engineering tool. However, when using these models to suggest new protein designs, one must deal with the vast combinatorial complexity of protein sequences. Here, we review how to use a sequence-to-function machine-learning surrogate model to select sequences for experimental measurement. First, we discuss how to select sequences through a single round of machine-learning optimization. Then, we discuss sequential optimization, where the goal is to discover optimized sequences and improve the model across multiple rounds of training, optimization, and experimental measurement.     

Multicategory individualized treatment regime using outcome weighted learning

Individualized treatment regimes (ITRs) aim to recommend treatments based on patient‐specific characteristics in order to maximize the expected clinical outcome. Outcome weighted learning approaches have been proposed for this optimization problem with primary focus on the binary treatment case. Many require assumptions of the outcome value or the randomization mechanism. In this paper, we propose a general framework for multicategory ITRs using generic surrogate risk. The proposed method accommodates the situations when the outcome takes negative value and/or when the propensity score is unknown. Theoretical results about Fisher consistency, excess risk, and risk consistency are established. In practice, we recommend using differentiable convex loss for computational optimization. We demonstrate the superiority of the proposed method under multinomial deviance risk to some existing methods by simulation and application on data from a clinical trial.     

Assessment of sources of uncertainty in macrophyte surveys and the consequences for river classification

The application of macrophytes in freshwater monitoring is still relatively limited and studies on their intercalibration and sources of variation are required. Therefore, the aim of the study was to compare selected indices and metrics based on macrophytes and to quantify their variability. During the STAR project, several aspects influencing uncertainty in estimation of the ecological quality of river were assessed. Results showed that several metrics based on the indicative value of plant species can be used in evaluation of the ecological status of rivers. Among estimated sources of variance in metric values the inter-surveyor differences had the lowest effect and slightly stronger were the influences of temporal variation (years and seasons) and shading. The impact of habitat modification was the most important factor. Analysis showed that some of macrophyte-based metrics (notably MTR and IBMR) are of sufficient precision in terms of sampling uncertainty, that they could be useful for estimating the ecological status of rivers in accordance with the aims of the Water Framework Directive.     

Efficient phenotypic sex classification of zebrafish using machine learning methods

Sex determination in zebrafish by manual approaches according to current guidelines relies on human observation. These guidelines for sex recognition have proven to be subjective and highly labor‐intensive. To address this problem, we present a methodology to automatically classify the phenotypic sex using two machine learning methods: Deep Convolutional Neural Networks (DCNNs) based on the whole fish appearance and Support Vector Machine (SVM) based on caudal fin coloration. Machine learning techniques in sex classification provide potential efficiency with the advantage of automatization and robustness in the prediction process. Furthermore, since developmental plasticity can be influenced by environmental conditions, we have investigated the impact of elevated water temperature during embryogenesis on sex and sex‐related differences in color intensity of adult zebrafish. The estimated color intensity based on SVM was then applied to detect the association between coloration and body weight and length. Phenotypic sex classifications using machine learning methods resulted in a high degree of association with the real sex in nontreated animals. In temperature‐induced animals, DCNNs reached a performance of 100%, whereas 20% of males were misclassified using SVM due to a lower color intensity. Furthermore, a positive association between color intensity and body weight and length was observed in males. Our study demonstrates that high ambient temperature leads to a lower color intensity in male animals and a positive association of male caudal fin coloration with body weight and length, which appears to play a significant role in sexual attraction. The software developed for sex classification in this study is readily applicable to other species with sex‐linked visible phenotypic differences.     

An ensemble learning integration of multiple CNN with improved vision transformer models for pest classification

Pests are the main threats to crop growth, and the precision classification of pests is conducive to formulating effective prevention and governance strategies. In response to the problems of low efficiency and inadaptability to the large-scale environment of existing pest classification methods, this paper proposes a new pest classification method based on a convolutional neural network (CNN) and an improved Vision Transformer model. First, the MMAlNet is designed to extract the characteristics of the identification object from different scales and finer granularity. Then, a classification model called DenseNet Vision Transformer (DNVT) combining a CNN and an improved vision transformer model is proposed. The proposed DNVT captures both long distance dependencies and local characteristic modelling capabilities, which can effectively improve pest classification accuracy. Finally, the ensemble learning algorithm is used to learn MMAlNet and DNVT classification forecasts for soft voting, further enhancing the classification accuracy of pests. The simulation experiment results on the D0 and IP102 datasets show that the proposed method attained a maximum classification of 99.89 and 74.20%, respectively, which is better than other state-of-the-art methods and has a high practical application value.     

The use of an unsupervised learning approach for characterizing latent behaviors in accelerometer data

The recent increase in data accuracy from high resolution accelerometers offers substantial potential for improved understanding and prediction of animal movements. However, current approaches used for analysing these multivariable datasets typically require existing knowledge of the behaviors of the animals to inform the behavioral classification process. These methods are thus not well‐suited for the many cases where limited knowledge of the different behaviors performed exist. Here, we introduce the use of an unsupervised learning algorithm. To illustrate the method's capability we analyse data collected using a combination of GPS and Accelerometers on two seabird species: razorbills (Alca torda) and common guillemots (Uria aalge). We applied the unsupervised learning algorithm Expectation Maximization to characterize latent behavioral states both above and below water at both individual and group level. The application of this flexible approach yielded significant new insights into the foraging strategies of the two study species, both above and below the surface of the water. In addition to general behavioral modes such as flying, floating, as well as descending and ascending phases within the water column, this approach allowed an exploration of previously unstudied and important behaviors such as searching and prey chasing/capture events. We propose that this unsupervised learning approach provides an ideal tool for the systematic analysis of such complex multivariable movement data that are increasingly being obtained with accelerometer tags across species. In particular, we recommend its application in cases where we have limited current knowledge of the behaviors performed and existing supervised learning approaches may have limited utility.     

Adaptive patient enrichment designs in therapeutic trials

The utility of clinical trial designs with adaptive patient enrichment is investigated in an adequate and well‐controlled trial setting. The overall treatment effect is the weighted average of the treatment effects in the mutually exclusive subsets of the originally intended entire study population. The adaptive enrichment approaches permit assessment of treatment effect that may be applicable to specific nested patient (sub)sets due to heterogeneous patient characteristics and/or differential response to treatment, e.g. a responsive patient subset versus a lack of beneficial patient subset, in all patient (sub)sets studied. The adaptive enrichment approaches considered include three adaptive design scenarios: (i) total sample size fixed and with futility stopping, (ii) sample size adaptation and futility stopping, and (iii) sample size adaptation without futility stopping. We show that regardless of whether the treatment effect eventually assessed is applicable to the originally studied patient population or only to the nested patient subsets; it is possible to devise an adaptive enrichment approach that statistically outperforms one‐size‐fits‐all fixed design approach and the fixed design with a pre‐specified multiple test procedure. We emphasize the need of additional studies to replicate the finding of a treatment effect in an enriched patient subset. The replication studies are likely to need fewer number of patients because of an identified treatment effect size that is larger than the diluted overall effect size. The adaptive designs, when applicable, are along the line of efficiency consideration in a drug development program.     

A hierarchical deep learning based approach for multi‐functional enzyme classification

Enzymes are critical proteins in every organism. They speed up essential chemical reactions, help fight diseases, and have a wide use in the pharmaceutical and manufacturing industries. Wet lab experiments to figure out an enzyme''s function are time consuming and expensive. Therefore, the need for computational approaches to address this problem are becoming necessary. Usually, an enzyme is extremely specific in performing its function. However, there exist enzymes that can perform multiple functions. A multi‐functional enzyme has vast potential as it reduces the need to discover/use different enzymes for different functions. We propose an approach to predict a multi‐functional enzyme''s function up to the most specific fourth level of the hierarchy of the Enzyme Commission (EC) number. Previous studies can only predict the function of the enzyme till level 1. Using a dataset of 2,583 multi‐functional enzymes, we achieved a hierarchical subset accuracy of 71.4% and a Macro F₁ Score of 96.1% at the fourth level. The robustness of the network was further tested on a multi‐functional isoforms dataset. Our method is broadly applicable and may be used to discover better enzymes. The web‐server can be freely accessed at http://hecnet.cbrlab.org/.     

Parallel reinforcement learning for weighted multi-criteria model with adaptive margin

Reinforcement learning (RL) for a linear family of tasks is described in this paper. The key of our discussion is nonlinearity of the optimal solution even if the task family is linear; we cannot obtain the optimal policy using a naive approach. Although an algorithm exists for calculating the equivalent result to Q-learning for each task simultaneously, it presents the problem of explosion of set sizes. We therefore introduce adaptive margins to overcome this difficulty.     

Random forest for gene selection and microarray data classification

A random forest method has been selected to perform both gene selection and classification of the microarray data. In this embedded method, the selection of smallest possible sets of genes with lowest error rates is the key factor in achieving highest classification accuracy. Hence, improved gene selection method using random forest has been proposed to obtain the smallest subset of genes as well as biggest subset of genes prior to classification. The option for biggest subset selection is done to assist researchers who intend to use the informative genes for further research. Enhanced random forest gene selection has performed better in terms of selecting the smallest subset as well as biggest subset of informative genes with lowest out of bag error rates through gene selection. Furthermore, the classification performed on the selected subset of genes using random forest has lead to lower prediction error rates compared to existing method and other similar available methods.     

Adaptive learning of host preference in a herbivorous arthropod

Although many publications deal with the effects of experience on behaviour, adaptive learning (i.e. behavioural change with experience resulting in improved reproductive success) is poorly documented. We present direct evidence that learning of host preference improves fitness in the herbivorous mite, Tetranychus urticae . Individual mites from two strains were repeatedly given a choice between two host plants, tomato and cucumber, and then subjected to a performance test on each. For both strains, food experience affected the subsequent choice: individual mites learned to prefer cucumber over tomato. The performance test showed this effect to be adaptive, as the food plant the mites learned to prefer (cucumber) allowed for increased oviposition, survival and development. These findings have important implications for the interpretation of the preference–performance relationship among herbivorous arthropods. The frequently reported absence of such a relationship may be due to experience-dependent preference and/or performance.     

基于无人机多源遥感数据的亚热带森林树种分类

树种多样性是生态学研究的重要内容,树木的种类和空间分布信息可有效服务于可持续森林管理。但在复杂林分条件下,获取高精度分类结果的难度大。而无人机遥感可获取局域超精细数据,为树种分类精度的提高提供了可能。基于可见光、高光谱、激光雷达等多源无人机遥感数据,探究其在亚热带林分条件下的树种分类潜力。研究发现:(1)随机森林分类器总体精度和各树种的F1分数最高,适合亚热带多树种的分类制图,其区分13种类别(8乔木,4草本)的总体精度为95.63%,Kappa系数为0.948;(2)多源数据的使用可以显著提高分类精度,全特征模型精度最高,且高光谱和激光雷达数据显著影响全特征模型分类精度,可见光纹理数据作用较小;(3)分类特征重要性从大到小排序为结构信息,植被指数,纹理信息,最小噪声变换分量。     

Learning to see the wood for the trees: machine learning,decision trees,and the classification of isolated theropod teeth

Taxonomic identification of fossils based on morphometric data traditionally relies on the use of standard linear models to classify such data. Machine learning and decision trees offer powerful alternative approaches to this problem but are not widely used in palaeontology. Here, we apply these techniques to published morphometric data of isolated theropod teeth in order to explore their utility in tackling taxonomic problems. We chose two published datasets consisting of 886 teeth from 14 taxa and 3020 teeth from 17 taxa, respectively, each with five morphometric variables per tooth. We also explored the effects that missing data have on the final classification accuracy. Our results suggest that machine learning and decision trees yield superior classification results over a wide range of data permutations, with decision trees achieving accuracies of 96% in classifying test data in some cases. Missing data or attempts to generate synthetic data to overcome missing data seriously degrade all classifiers predictive accuracy. The results of our analyses also indicate that using ensemble classifiers combining different classification techniques and the examination of posterior probabilities is a useful aid in checking final class assignments. The application of such techniques to isolated theropod teeth demonstrate that simple morphometric data can be used to yield statistically robust taxonomic classifications and that lower classification accuracy is more likely to reflect preservational limitations of the data or poor application of the methods.     

Adaptive learning in arthropods: spider mites learn to distinguish food quality

Many herbivorous arthropods have been shown to possess learning capabilities, yet fitness effects of learning are largely unknown. In this paper, we test whether two-spotted spider mites (Tetranychus urticae) learn to distinguish food quality in choice tests, and whether this results in fitness benefits. Food consisted of cucumber plants with one of three degrees of feeding damage: undamaged (no mites), mildly damaged (infested by a mite strain adapted to tomato) and heavily damaged (infested by a mite strain adapted to cucumber). Mites were subjected to one choice test in a greenhouse and three sequential choice tests on leaf disks. Thereafter, individual mite performance was measured as oviposition rate over four days. In the course of the three small-scale choice tests, preference shifted towards less damaged food. The performance tests showed that learning was adaptive: mites learned to prefer the food type that yielded the higher oviposition rate. Interestingly, innate preferences in the greenhouse tests were close to those shown after learning in the small-scale tests. Given that both strains of mites had not experienced cucumber for several years, we hypothesize that the preference in the greenhouse was due to avoidance of mite odours rather than odours of damaged plants. Through its effect on foraging behaviour, adaptive learning may promote the evolution of host plant specialization in herbivorous arthropods. This revised version was published online in July 2006 with corrections to the Cover Date.