Background: The massive amount of biomedical data accumulated in the past decades can be utilized for diagnosing disease. Objective: However, the high dimensionality, small sample sizes, and irrelevant features of data often have a negative influence on the accuracy and speed of disease prediction. Some existing machine learning models cannot capture the patterns on these datasets accurately without utilizing feature selection. Methods: Filter and wrapper are two prevailing feature selection methods. The filter method is fast but has low prediction accuracy, while the latter can obtain high accuracy but has a formidable computation cost. Given the drawbacks of using filter or wrapper individually, a novel feature selection method, called MRMR-EFPATS, is proposed, which hybridizes filter method Minimum Redundancy Maximum Relevance (MRMR) and wrapper method based on an improved Flower Pollination Algorithm (FPA). First, MRMR is employed to rank and screen out some important features quickly. These features are further chosen for individual populations following the wrapper method for faster convergence and less computational time. Then, due to its efficiency and flexibility, FPA is adopted to further discover an optimal feature subset. Result: FPA still has some drawbacks, such as slow convergence rate, inadequacy in terms of searching new solutions, and tends to be trapped in local optima. In our work, an elite strategy is adopted to improve the convergence speed of the FPA. Tabu search and Adaptive Gaussian Mutation are employed to improve the search capability of FPA and escape from local optima. Here, the KNN classifier with the 5-fold-CV is utilized to evaluate the classification accuracy. Conclusion: Extensive experimental results on six public high dimensional biomedical datasets show that the proposed MRMR-EFPATS has achieved superior performance compared to other state-of-theart methods.

A discontinuous Galerkin time-domain electric field integral equation method based on a marching-on-in-degree (MOD) scheme for simulating transient responses of perfect electric conductor objects is proposed in this letter. With the help of half-Rao-Wilton-Glisson spatial basis functions, we can simulate transient scattering responses of three-dimensional structures with either conformal or nonconformal meshes. The weighted Laguerre polynomials are chosen as temporal basis functions and implemented for the MOD scheme. Under such circumstances, the developed algorithm eliminates late-time instabilities and the limitation of Courant-Friedrichs-Lewy criteria. In addition, we adopt a fast-Fourier-transform-based blocking scheme to accelerate the temporal convolution of the MOD procedure. Some numerical examples are given to validate its stability and accuracy.

The current study investigated the facilitatory effects of High Variability Phonetic Training (HVPT) in second language (L2) categorical perception (CP) of Mandarin lexical tones. It also explored whether and how individual differences in auditory processing predicted gains from such training. The participants were 32 native English-speaking adults aged over 60 years who were learning Mandarin Chinese as their L2. They were randomly divided into the HVPT group (HG) ( n = 16) and the control group (CG) ( n = 16). Their L2 CP performance was assessed through an identification task and discrimination task before training, immediately after training, and two months later. Auditory processing tests were also conducted to measure the participants’ ability to encode spectral and temporal details of sounds. Linear mixed-effects (LME) models showed that, compared to the CG, the HG exhibited a more pronounced improvement in tonal categorization. Furthermore, regression analysis confirmed that individual differences in perceptual acuity significantly predicted gains from training in L2 CP of Mandarin lexical tones.

The present study investigated the differential effects of explicit corrections, meta-linguistic corrective feedback (CF), and analogy-based CF on L2 learners' acquisition of English third-person singular form -s and whether and how individual differences in working memory (WM) mediate such effects. One hundred secondary school English-as-a-foreign-language (EFL) learners at a junior middle school in inland China were randomly assigned to the explicit correction group (EG), the meta-linguistic CF group (MG), the analogy-based CF group (AG), and the control group (CG). Learners performed both an information-gap activity and a picture-description activity where their errors on target structure were treated according to their group assignment. The Untimed Grammatical Judgement Test (UGJT) and the Elicited Oral Production Test (EOPT) were used to measure learners' resulting performance. Learners' WM was measured with operation span test. Results revealed that (1) compared to the control group, all the CF groups significantly improved their performance of English third-person singular form -s over time; (2) explicit corrections and meta-linguistic CF displayed superior advantages over analogy-based CF on the immediate posttest. However, the three CF groups demonstrated no significant difference in their performance of English third-person singular form -s on the delayed posttest; (3) WM was only able to predict the effects of analogy-based CF but not explicit corrections and meta-linguistic CF; and (4) analogy-based CF was more favorable to learners with higher WM who can regulate their limited attentional resources more efficiently, whereas explicit corrections and meta-linguistic CF equalize learning opportunities for all learners with different levels of WM. The findings of this study suggest optimal, profile-matched pedagogical options for L2 learning through identifying CF conditions that cater to the needs of young learners with different levels of WM.

This study used the categorical perception (CP) paradigm, a fine-grained perceptual method, to investigate the perceptual performance of lexical tones in Chinese people with post-stroke aphasia (PWA). Twenty patients with post-stroke aphasia (10 Broca’s and 10 Wernicke’s) and ten neurologically intact age-matched control participants were recruited to complete both identification and discrimination tasks of the Mandarin Tone 1–2 continuum. In addition, all participants completed tests on their auditory comprehension ability and working memory. The results showed that both Broca’s and Wernicke’s patients exhibited reduced sensitivity to within-category and between-category information but preserved CP of lexical tones. The degree of CP of lexical tones related to working memory in aphasic patients. Furthermore, lower-level acoustic processing underpinned higher-level phonological processing on the CP of lexical tones since both patient groups’ unbalanced pitch processing ability extended to their CP of lexical tones. These findings are significant for researchers and clinicians in speech-language rehabilitation, clinical psychology, and cognitive communication.

Developing adequate second language (L2) speech has increasingly become one of the important goals in today’s globalized world. The current study set out to investigate how the provision of recasts can facilitate English vowels /i/-/ɪ/ learning by Chinese native speakers and how such learning gains resulting from recasts can be tied to individual differences in auditory processing. An experimental group of thirty young adults was engaged in ten treatment sessions of recasts lasting ninety minutes each to notice and practice English vowels /i/-/ɪ/. Their learning gains were tested through a forced-choice identification task, a controlled word-reading task, and a spontaneous picture-narrative task before and after receiving recasts. A control group of the same number of participants was also included to receive comparable sessions without receiving any recasts to check for test-retest effects. Quantitative analysis showed that recasts significantly helped Chinese native speakers facilitate their learning of English vowels /i/-/ɪ/ regardless of time and lexical context. Such learning gains resulting from recasts were significantly correlated with specific constructs of auditory processing. Namely, perceptual acuity was significantly correlated with gains from recasts in identifying L2 vowels, and audio-motor integration with gains from recasts in controlled/spontaneous production of L2 vowels. The findings suggest that recasts can serve as ideal instructional techniques of L2 vowel learning for Chinese native speakers and that the auditory processing profiles of individuals can suggest how they can benefit the most from recasts.