Search

In most traditional machine learning algorithms, the training and testing datasets have identical distributions and feature spaces. However, these assumptions have not held in many real applications. Although transfer learning methods have been invented to fill this gap, they introduce new challenges as negative transfers (NTs). Most previous research considered NT a significant problem, but they pay less attention to solving it. This study will propose a transductive learning algorithm based on cellular learning automata (CLA) to alleviate the NT issue. Two famous learning automata (LA) entitled estimators are applied as estimator CLA in the proposed algorithms. A couple of new decision criteria called merit and and attitude parameters are introduced to CLA to limit NT. The proposed algorithms are applied to standard LA environments. The experiments show that the proposed algorithm leads to higher accuracy and less NT results.

Some researchers have introduced transfer learning mechanisms to multiagent reinforcement learning (MARL). However, the existing works devoted to cross-task transfer for multiagent systems were designed just for homogeneous agents or similar domains. This work proposes an all-purpose cross-transfer method, called multiagent lateral transfer (MALT), assisting MARL with alleviating the training burden. We discuss several challenges in developing an all-purpose multiagent cross-task transfer learning method and provide a feasible way of reusing knowledge for MARL. In the developed method, we take features as the transfer object rather than policies or experiences, inspired by the progressive network. To achieve more efficient transfer, we assign pretrained policy networks for agents based on clustering, while an attention module is introduced to enhance the transfer framework. The proposed method has no strict requirements for the source task and target task. Compared with the existing works, our method can transfer knowledge among heterogeneous agents and also avoid negative transfer in the case of fully different tasks. As far as we know, this article is the first work denoted to all-purpose cross-task transfer for MARL. Several experiments in various scenarios have been conducted to compare the performance of the proposed method with baselines. The results demonstrate that the method is sufficiently flexible for most settings, including cooperative, competitive, homogeneous, and heterogeneous configurations.

Reinforcement learning (RL) plays an essential role in the field of artificial intelligence but suffers from data inefficiency and model-shift issues. One possible solution to deal with such issues is to exploit transfer learning. However, interpretability problems and negative transfer may occur without explainable models. In this article, we define Relation Transfer as explainable and transferable learning based on graphical model representations, inferring the skeleton and relations among variables in a causal view and generalizing to the target domain. The proposed algorithm consists of the following three steps. First, we leverage a suitable casual discovery method to identify the causal graph based on the augmented source domain data. After that, we make inferences on the target model based on the prior causal knowledge. Finally, offline RL training on the target model is utilized as prior knowledge to improve the policy training in the target domain. The proposed method can answer the question of what to transfer and realize zero-shot transfer across related domains in a principled way. To demonstrate the robustness of the proposed framework, we conduct experiments on four classical control problems as well as one simulation to the real-world application. Experimental results on both continuous and discrete cases demonstrate the efficacy of the proposed method.

Existing transfer learning methods that focus on problems in stationary environments are not usually applicable to dynamic environments, where concept drift may occur. To the best of our knowledge, the concept drift-tolerant transfer learning (CDTL), whose major challenge is the need to adapt the target model and knowledge of source domains to the changing environments, has yet to be well explored in the literature. This article, therefore, proposes a hybrid ensemble approach to deal with the CDTL problem provided that data in the target domain are generated in a streaming chunk-by-chunk manner from nonstationary environments. At each time step, a class-wise weighted ensemble is presented to adapt the model of target domains to new environments. It assigns a weight vector for each classifier generated from the previous data chunks to allow each class of the current data leveraging historical knowledge independently. Then, a domain-wise weighted ensemble is introduced to combine the source and target models to select useful knowledge of each domain. The source models are updated with the source instances performed by the proposed adaptive weighted CORrelation ALignment (AW-CORAL). AW-CORAL iteratively minimizes domain discrepancy meanwhile decreases the effect of unrelated source instances. In this way, positive knowledge of source domains can be potentially promoted while negative knowledge is reduced. Empirical studies on synthetic and real benchmark data sets demonstrate the effectiveness of the proposed algorithm.

Transfer learning algorithm can provide a framework to utilize the previous knowledge to train fuzzy neural network (FNN). However, the performance of TL-based FNN will be destroyed by the knowledge over-fitting problem in the learning process. To solve this problem, an interactive transfer learning (ITL) algorithm, which can alleviate the negative transfer among different domains to improve the learning performance of FNN, is designed and analyzed in this paper. This ITL-assisted FNN (ITL-FNN) contains the following advantages. Frist, a knowledge filter algorithm is developed to reconstruct the knowledge in source scene by balancing the matching accuracy and diversity. Then, the knowledge from source scene can fit the instance of target scene with suitable accuracy. Second, a self-balancing mechanism is designed to balance the driven information between the source and target scenes. Then, the knowledge can be refitted to reduce the useless information. Third, a structural competition algorithm is proposed to adjust the knowledge of FNN. Then, the proposed ITL-FNN can achieve compact structure to improve the generalization performance. Finally, some benchmark problems and industrial applications are provided to demonstrate the merits of ITL-FNN.

The research of semantic segmentation based on unsupervised domain adaptation greatly alleviates the high-cost bottleneck of manual annotation in deep learning. Inevitably domain gap limits the ability of target domain to learn knowledge from source domain. Previous domain alignment strategies aim to explore maximum domain-invariant space to enlarge knowledge that target domain learns. They based on the assumption that images are symmetrical in categories covering all predefined categories. However, in practical training, there is class asymmetry in image samples inter domains. This will lead to outlier categories and negative transfer, because it lacks the guidance of the corresponding category of the source domain. To tackle this issue, we propose a partial domain adaptive method on semantic segmentation to guide target model to learn category-level knowledge selectively. PAM (Partial Adaptive Map) module is introduced to motivate the target model to obtain more knowledge from non-outliers and less knowledge from outliers to avoid negative transfer. We further analyze the effectiveness of this method from the perspective of JS divergence. Our method has achieved significant improvement without additional discriminators. Experiments on general datasets GTA5 and SYNTHIA can compare with SOTA methods.

Evolutionary multi-task optimization (EMTO) has recently attracted widespread attention in the evolutionary computation community, which solves two or more tasks simultaneously to improve the convergence characteristics of tasks when individually optimized. Effective knowledge between tasks is transferred by taking advantage of the parallelism of population-based search. Without any prior knowledge about tasks, it is a challenging problem of how to adaptively transfer effective knowledge between tasks and reduce the impact of negative transfer in EMTO. However, these two issues are rarely studied simultaneously in the existing literature. Besides, in complex many-task environments, the potential relationships among individuals from highly diverse populations associated with tasks directly determine the effectiveness of cross-task knowledge transfer. Keeping those in mind, we propose a multi-task evolutionary algorithm based on anomaly detection (MTEA-AD). Specifically, each task is assigned a population and an anomaly detection model. Each anomaly detection model is used to learn the relationship among individuals between the current task and the other tasks online. Individuals that may carry negative knowledge are identified as outliers. And candidate transferred individuals identified by the anomaly detection model are selected to assist the current task, which may carry common knowledge across the current task and other tasks. Furthermore, to realize the adaptive control of the degree of knowledge transfer, the successfully transferred individuals that survive to the next generation through the elitism are used to update the anomaly detection parameter. The fair competition between offspring and candidate transferred individuals can effectively reduce the risk of negative transfer. Finally, the empirical studies on a series of synthetic benchmarks and a practical study are conducted to verify the effectiveness of MTEA-AD. The experimental results demonstrate that our proposal can adaptively adjust the degree of knowledge transfer through the anomaly detection model to achieve highly competitive performance compared to several state-of-the-art EMTO methods.

Multi-task optimization aims to solve two or more optimization tasks simultaneously by leveraging inter-task knowledge transfer. However, as the number of tasks increases to the extent of many-task optimization, the knowledge transfer between tasks encounters more uncertainty and challenges, thereby resulting in degradation of optimization performance. To give full play to the many-task optimization framework and minimize the potential negative transfer, this paper proposes an evolutionary many-task optimization algorithm based on a multi-source knowledge transfer mechanism namely EMaTO-MKT. Particularly, in each iteration, EMaTO-MKT determines the probability of using knowledge transfer adaptively according to the evolution experience, and balances the self-evolution within each task and the knowledge transfer among tasks. To perform knowledge transfer, EMaTO-MKT selects multiple highly similar tasks in term of maximum mean discrepancy as the learning sources for each task. Afterward, a knowledge transfer strategy based on local distribution estimation is applied to enable the learning from multiple sources. Compared with the other state-of-the-art evolutionary many-task algorithms on benchmark test suites, EMaTO-MKT shows competitiveness in solving many-task optimization problems.

Particle swarm optimization algorithm has become a promising approach in solving multi-task optimization (MTO) problems since it can transfer knowledge with easy implementation and high searching efficiency. However, in the process of knowledge transfer, negative transfer is common because it is difficult to evaluate whether knowledge is effective for population evolution. Therefore, how to obtain and transfer the effective knowledge to curb the negative transfer is a challenging problem in MTO. To deal with this problem, a self-adjusting multi-task particle swarm optimization (SA-MTPSO) algorithm is designed to improve the convergence performance in this paper. First, a knowledge estimation metric, combining the decision space knowledge and the target space knowledge for each task, is designed to describe the effectiveness of knowledge. Then, the effective knowledge is obtained to promote the knowledge transfer process. Second, a self-adjusting knowledge transfer mechanism, based on the effective knowledge and the self-adjusting transfer method, is developed to achieve effective knowledge transfer. Then, the ineffective knowledge is removed to solve the negative transfer problem. Third, the convergence analysis is given to guarantee the effectiveness of the SA-MTPSO algorithm theoretically. Finally, the proposed algorithm is compared with some existing MTO algorithms. The results show that the performance of the proposed algorithm is superior to most algorithms on negative transfer suppression and convergence.

Over the past decade, domain adaptation (DA) algorithms have been proposed to address domain gap problems as they do not need any interpretation in the target domain. However, most existing efforts focus on scenarios with only one source domain and one target domain. In this article, we explore the scenario with one source domain and mixed multiple target domains for remote sensing applications and propose a new algorithm, named the two-stage adaptation network (TSAN). First, we utilize the adversarial learning approach to confuse the classifier to discriminate between the source domain and the whole mixed-multiple-target domain. Second, we adopt self-supervised learning to divide the mixed-multiple-target domain with automated generation of ``pseudo''-domain labels, which guides our network to learn intrinsic features of multiple target domains. Finally, these two steps are combined as an iterative procedure. We integrate a test dataset that includes five remote sensing datasets and ten classes. Our method achieves an average accuracy of 63.25% and 73.68% with two typical backbones, considerably outperforming other DA methods with an average accuracy improvement of 4.84%-20.19% and 9.06%-17.04%, respectively. Furthermore, we identify the negative transfer effect in existing mainstream DA methods in remote sensing image classification with multiple different domains.

In recent years, many efficient evolutionary multitasking (EMT) algorithms have been proposed to solve multi-objective multi-task optimization problems. However, EMT algorithms often face negative transfer problems. In this paper, a novel multi-objective evolution strategy, called CT-EMT-MOES, is proposed based on a cultural transmission theory for solving multi-objective multitask optimization problems. First, two evolutionary operators inspired by cultural transmission theory are proposed. The elite-guided variation strategy can transfer the information from the current Pareto front to all individuals and guide the population to quickly converge. The horizontal cultural transmission strategy can efficiently transfer information from the source task to the target task. Second, to solve the negative transfer problem, an adaptive information transfer strategy is proposed to adaptively adjust the probability of an information transfer. Third, the proposed algorithm can gain a Pareto front with good convergence and diversity by utilizing a smaller population size and fewer computing resources. As a result, the proposed algorithm can effectively utilize the implicit similarity and complementarity between simultaneous optimized tasks to improve the overall convergence efficiency and reduce a negative transfer. Finally, comprehensive experimental results show that the proposed algorithm can achieve a better performance compared with previous state-of-the-art multi-objective EMT algorithms.

The paper deals with the problem of poor acquisition of English nominal compounds in EFL university students who are native speakers of Bosnian/Croatian/Serbian (BCS). The problem was tackled by a brief comparison of noun morphology and declension suffixes in English and BCS, the introduction of negative transfer in the process of foreign language acquisition, and the analysis of BCS translation equivalents. The discussion revealed that the productivity of compounding largely depends on the complexity of noun morphology. Put simply, the more complex noun morphology is, the less productive compounding will be. In addition, the analysis of the translation equivalents also highlighted that a vast majority of problems related to poor acquisition of English nominal compounds stems from the fact that EFL learners usually translate L1 phrases word-forword into L2. Finally, it was pointed out that teaching English nominal compounds to EFL learners could be improved by tailored-made instructions provided in L1.

Errors, defined as mistaken application of linguistic theory done by language learners, are generally seen as negative element in language learning. However, several researchers believed that error analysis can be used to understand how students process a target language. Understanding this will give insights on which areas of language the students find it difficult. Therefore, this study was conducted for this purpose. Descriptive qualitative methodology was employed to examine types of writing errors that are related to inflectional affixation performed by 8 freshman students. The errors that were identified in the students’ written language were described in linguistic and surface category. Authoritative interpretation was conducted to investigate the cause of error occurrence through interview. The study finds that the students still face difficulty in subject-verb agreement, plurals, and past participle. Furthermore, it was also found that the dominant cause of the error occurrence is interlingual factors, negative transfer from their first language.

As an effective learning paradigm against insufficient training samples, Multi-Task Learning (MTL) encourages knowledge sharing across multiple related tasks so as to improve the overall performance. In MTL, a major challenge springs from the phenomenon that sharing the knowledge with dissimilar and hard tasks, known as negative transfer, often results in a worsened performance. Though a substantial amount of studies have been carried out against the negative transfer, most of the existing methods only model the transfer relationship as task correlations, with the transfer across features and tasks left unconsidered. Different from the existing methods, our goal is to alleviate negative transfer collaboratively across features and tasks. To this end, we propose a novel multi-task learning method called Task-Feature Collaborative Learning (TFCL). Specifically, we first propose a base model with a heterogeneous block-diagonal structure regularizer to leverage the collaborative grouping of features and tasks and suppressing inter-group knowledge sharing. We then propose an optimization method for the model. Extensive theoretical analysis shows that our proposed method has the following benefits: (a) it enjoys the global convergence property and (b) it provides a block-diagonal structure recovery guarantee. As a practical extension, we extend the base model by allowing overlapping features and differentiating the hard tasks. We further apply it to the personalized attribute prediction problem with fine-grained modeling of user behaviors. Finally, experimental results on both simulated dataset and real-world datasets demonstrate the effectiveness of our proposed method, Comment: To appear in T-PAMI

Negative transfer of mother tongue is a major difficulty that every second language learner has to overcome, especially for those students whose target language is quite different from their mother tongue. With the in-depth development of pluralism in the world, people pay more and more attention to the learning of various languages, especially English. In this situation, the concerns of language researchers on the negative transfer of mother tongue are inevitably boosting. Given this, based on the pertinent theories of the negative language transfer, this study aims to explore the impact of negative language transfer on Chinese college students’ English learning through a questionnaire poll with some students of Zhejiang Yuexiu University as the research objects. The results of the study are mainly shown in two aspects: the impact of the negative transfer of mother tongue on college students’ oral learning and college students’ writing learning. At the end of the article, the author puts forward the cultural differences between China and the Western countries that lead to the negative transfer of mother tongue, and some effective methods to help Chinese college English learners reduce the negative transfer of mother tongue in their process of English learning.

Using intermediate and advanced Chinese students as the research subjects, HSK corpus and questionnaire surveys were used to explore the error types in the construction of “X??(Y)+?+Z.” The types of errors can be divided into component errors, relationship errors between components, and collocation errors between components and sentences. The main reasons for the errors are the particularity of the construction, the immature development of construction teaching, the negative transfer of students’ mother tongue, as well as the generalization and avoidance of the target language.

This article analyzes errors with regard to the types and causes of errors commonly committed by Chinese EFL (English as a foreign language) learners. It covers the concepts and theories concerning error analysis, synthesizes previous research, as well as identifies the flaws and gaps found in those literatures. This article provides several teaching suggestions based on previous analyses, which have found that the misuse of verbs and articles are the most common errors in English writing among Chinese students. Based on this, most studies attributed those errors to the negative transfer of Chinese due to the thought patterns and disparities between Chinese and English.

Remaining useful life (RUL) prediction can effectively avoid unexpected mechanical breakdowns, thus improving operational reliability. However, the distribution discrepancy caused by different working conditions may lead to deterioration in the prognostic task of machinery. Inspired by the idea of transfer learning, a novel intelligent approach based on dynamic domain adaptation (DDA) is proposed for the machinery RUL prediction of multiple working conditions in this paper. At first, reverse validation technology is utilized to select appropriate source samples to construct the training dataset. Then two dynamic domain adaptation networks are trained to extract domain invariant degradation feature and predict RUL, namely dynamic distribution adaptation network and dynamic adversarial adaptation network. In the dynamic domain adaptation network, the fuzzy set theory is employed to calculate conditional distribution discrepancy loss, and the dynamic adaptive factor is introduced to dynamically adjust the distribution weights. Finally, the proposed method is proved to be effective through two run-to-failure bearing datasets. Related experimental results indicate that, compared with other related RUL prediction methods, the DDA-based prognostic method not only achieves better prediction performance, but also avoids the influence of negative transfer and distribution weight fluctuation.

Among different recommendation techniques, collaborative filtering usually suffers from limited performance due to the data sparsity problem. Transfer learning presents an unprecedented opportunity to alleviate this issue by transferring rating patterns from the source domain to the target domain. However, an issue occurs in some situations when the source and target domains have unrelated parts to each other, such that a negative transfer may reduce the learning performance in the target domain. To address this issue, in this article, we present a novel model called cluster rating bias restoration to divide the learning process into two stages. First, we employ traditional transfer learning methods to adapt and reconstruct the target rating matrix. Second, we propose a cluster rating bias restoration framework specifically to minimize the effect of the negative transfer from the source domain and to restore the context-information-based rating bias trend in the target domain. Experiments on real-world datasets show that our proposed model outperforms the state-of-the-art alternatives.

Motivation Large-scale cancer omics studies have highlighted the diversity of patient molecular profiles and the importance of leveraging this information to deliver the right drug to the right patient at the right time. Key challenges in learning predictive models for this include the high-dimensionality of omics data and heterogeneity in biological and clinical factors affecting patient response. The use of multi-task learning techniques has been widely explored to address dataset limitations for in vitro drug response models, while domain adaptation (DA) has been employed to extend them to predict in vivo response. In both of these transfer learning settings, noisy data for some tasks (or domains) can substantially reduce the performance for others compared to single-task (domain) learners, i.e. lead to negative transfer (NT). Results We describe a novel multi-task unsupervised DA method (TUGDA) that addresses these limitations in a unified framework by quantifying uncertainty in predictors and weighting their influence on shared feature representations. TUGDA’s ability to rely more on predictors with low-uncertainty allowed it to notably reduce cases of NT for in vitro models (94% overall) compared to state-of-the-art methods. For DA to in vivo settings, TUGDA improved over previous methods for patient-derived xenografts (9 out of 14 drugs) as well as patient datasets (significant associations in 9 out of 22 drugs). TUGDA’s ability to avoid NT thus provides a key capability as we try to integrate diverse drug-response datasets to build consistent predictive models with in vivo utility. Availabilityand implementation https://github.com/CSB5/TUGDA. Supplementary information Supplementary data are available at Bioinformatics online.

Experimental results from the paper titled "Negative Transfer in Cross Project Defect Prediction: Effect of Domain Divergence"

This study examined whether Korean learners of English attained native-like performance in English focus prosody by conducting production and perception experiments using digit strings. Language learners were classified into advanced-, intermediate-, and low-level groups according to their proficiency and compared with native speakers. Native speakers’ focus prosody was clearly prominent in the focus positions, and their post-focus positions were considerably compressed. Their focused digits were easy to detect, resulting in a 97% identification rate. Although advanced-level speakers produced acoustic cues quite similar to those of native speakers, their post-focus production did not resemble that of native speakers. Their identification rate was 81%, 16% lower than that of native speakers. Neither intermediate- nor low-level speakers’ focus-cueing changes were distinguished whatsoever in the focus and post-focus positions. Their identification rates were just over 10%, similar to the level of chance in a 10-digit string, implying that their focus productions were not sufficiently salient to be recognized in the experiment. The results suggest that second language acquisition is hindered by a negative transfer between English and Korean. The acquisition of second language focus prosody proceeds slowly; second language learners approach native-like proficiency once they become advanced.

Despite recent remarkable progress, person re-identification (ReID) still suffers from a shortage of annotated training data. To deal with this problem, there has been a boost of interest in developing various data augmentation methods. In this paper, we are devoted to developing an end-to-end joint representation learning framework for the ReID task on the basis of a novel data augmentation strategy. Specifically, we regard the original training dataset as a source domain and generate the counterpart augmented domains through image channel shuffling. Accordingly, we design a symmetric classification network for ReID learning. By investigating the domain-level and identity-level relationship between domains, we use the idea of negative transfer and structural consistency to optimize the network for learning discriminative feature embeddings. Comprehensive experiments on some benchmark datasets demonstrate the effectiveness and robustness of our proposed approach. Source code is released at: https://github.com/flychen321/negative_transfer_reid .

While there is some research into the role of the target language community in understanding negative transfer and refusals (e.g., Beebe, Takahashi, Uliss-Weltz,1990; Ikoma & Shimura, 1993; Takahashi & Beebe, 1987; Yamagashira, 2001), most research has focused on determining the extent to which L2 proficiency influences negative transfer without close attention to status of the interlocuter or face-saving strategies (e.g., Félix-Brasdefer, 2004; Chang, 2009, Jiang, 2015). In response to this gap in the research, this study reports on findings from an investigation into the negative transfer of refusals among 59 students studying in Japan and the USA. Findings demonstrated that learning environments play a role in determining the frequency of negative transfers and that negative transfer was sensitive to status. Content-related findings suggest that the production of face-saving refusals is sensitive to the learning environment as well. Implications for pragmatic research comparing L1 and L2 environments are offered.

Gaussian mixture model (GMM) is a well-known model-based approach for data clustering. However, when the data samples are insufficient, the classical GMM-based clustering algorithms are not effective anymore. Referring to the idea of transfer clustering methods, this paper proposes a general transfer GMM-based clustering framework, which employs the important knowledge extracted from some known source domain to guide and improve the clustering on the target domain with small-scale data. Specifically, three traditional GMM-based clustering approaches are extended to the corresponding transfer clustering versions. Furthermore, to avoid the negative transfer problem, maximum mean discrepancy (MMD) is introduced to search the most matched source domain to provide more positive guidance for data clustering on the target domain. Experiments on synthetic and real-world datasets demonstrate the efficiency of the presented framework compared with several existing transfer clustering algorithms.

This study focuses on the interference of L1 grammatical rules in the writing of L2 with specific reference to the interference of Tamil (L1) in English (L2). Though many research have been done in mother tongue interference, very few research have been conducted in terms of Tamil versus English. This study seeks to find out the components of the L1 (Tamil language) grammar that the students of Tamil schools use interchangeably in their daily writing of L2, namely English. This study will follow a mixture of qualitative and quantitative survey research design, the purpose of which is to find out the influence of mother tongue linguistic items in the writing of English (L2), among the standard 4 students of a selected school. The findings of the study clearly indicated that mother tongue and the national language interfered (negative transfer) in the writing of L2 among Tamil school students. Apart from the interference of L1, the research also found that there is a heavy influence of the national language (Malay) in the writing of students' L2.

Dynamic multiobjective optimization problems (DMOPs) are optimization problems with multiple conflicting optimization objectives, and these objectives change over time. Transfer learning-based approaches have been proven to be promising; however, a slow solving speed is one of the main obstacles preventing such methods from solving real-world problems. One of the reasons for the slow running speed is that low-quality individuals occupy a large amount of computing resources, and these individuals may lead to negative transfer. Combining high-quality individuals, such as knee points, with transfer learning is a feasible solution to this problem. However, the problem with this idea is that the number of high-quality individuals is often very small, so it is difficult to acquire substantial improvements using conventional transfer learning methods. In this article, we propose a knee point-based transfer learning method, called KT-DMOEA, for solving DMOPs. In the proposed method, a trend prediction model (TPM) is developed for producing the estimated knee points. Then, an imbalance transfer learning method is proposed to generate a high-quality initial population by using these estimated knee points. The advantage of this approach is that the seamless integration of a small number of high-quality individuals and the imbalance transfer learning technique can greatly improve the computational efficiency while maintaining the quality of the solution. The experimental results and performance comparisons with some chosen state-of-the-art algorithms demonstrate that the proposed design is capable of significantly improving the performance of dynamic optimization.

Do handwriting skills transfer when a child writes in two different scripts, such as the Latin and Cyrillic alphabets? Are our measures of handwriting skills intrinsically bound to one alphabet or will a child who faces handwriting difficulties in one script experience similar difficulties in the other script? To answer these questions, 190 children from grades 1–4 were asked to copy a short text using both the Cyrillic and Latin alphabets on a digital tablet. A recent change of policy in Kazakhstan gave us an opportunity to measure transfer, as the Latin-based Kazakh alphabet has not yet been introduced. Therefore, pupils in grade 1 had a 6-months experience in Cyrillic, and pupils in grades 2, 3, and 4 had 1.5, 2.5, and 3.5 years of experience in Cyrillic, respectively. This unique situation created a quasi-experimental situation that allowed us to measure the influence of the number of years spent practicing Cyrillic on the quality of handwriting in the Latin alphabet. The results showed that some of the differences between the two scripts were constant across all grades. These differences thus reflect the intrinsic differences in the handwriting dynamics between the two alphabets. For instance, several features related to the pen pressure on the tablet are quite different. Other features, however, revealed decreasing differences between the two scripts across grades. While we found that the quality of Cyrillic writing increased from grades 1–4, due to increased practice, we also found that the quality of the Latin writing increased as well, despite the fact that all of the pupils had the same absence of experience in writing in Latin. We can therefore interpret this improvement in Latin script as an indicator of the transfer of fine motor control skills from Cyrillic to Latin. This result is especially surprising given that one could instead hypothesize a negative transfer, i.e., that the finger controls automated for one alphabet would interfere with those required by the other alphabet. One interesting side-effect of these findings is that the algorithms that we developed for the diagnosis of handwriting difficulties among French-speaking children could be relevant for other alphabets, paving the way for the creation of a cross-lingual model for the detection of handwriting difficulties.

Research studies concerning pragmatic acquisition by students have established that pragmatic nonconformity or deviation in the execution of speech acts, among which is "condolence", is primarily due to the students' incompetence to identify the proper meaning and to handle the proper form. In Iraqi context, most Iraqi EFL university students are found to inapproriately use illocutionary acts, as they digress from the sociocultural standards and pragmalinguistic rules used for making condolence. When producing circumstances comprising the use of the expressive condolence, students' speeches and responses do not conform to those of native speakers. Their replies seem awkward or refer to certain spoken situations other than those required in the communicative process or interaction. Thus, the study intends to identify Iraqi EFL students' deviation in the performance of the expressive condolence in order to get at the reasons behind this unconventionality. To achieve this aim, fifty Iraqi EFL college students of the University of Baghdad , College of Languages, Department of English, have been chosen to give ( data) replies to a test containing five situtations involving the production of condolence. Responses to the test have shown that negative transfer of L1 socio-pragmatic knowledge and lack of pragma-linguistic rules are the main characteristics that EFL Iraqi students resort to when handling condolence. Socio-pragmatic deviation occurs because they are linguistically unconscious of the conventions and means used in the target language; therefore, their performance is a non-native one. Further, lack of syntactic and semantic knowledge is an influential factor in producing certain spoken situations of condoling.

Mobile learning has brought significant changes in English language teaching and learning. Fun Dubbing, as one of the Apps to learn English pronunciation, has drawn extensive attention among students. This paper aims to analyze the errors that junior high school students make in Fun Dubbing based on Error Analysis. The results showed that negative transfer of Hanyu Pinyin, wrong distinction between short vowels and long vowels and lack of knowledge about confusing phonemes are three major causes of errors. In addition, some suggestions are provided for English language learners.

Convolutional neural networks have demonstrated state-of-the-art performance in image classification and various other computer vision tasks. Plant disease detection is an important area of deep learning which has been addressed by many recent methods. However, there is a dire need to optimize these solutions for resource-constrained portable devices such as smartphones. This is a challenging problem because deep learning models are resource extensive in nature. This paper proposes an efficient method to systematically classify plant disease symptoms using convolutional neural networks. These networks are memory efficient and when coupled with the proposed training configuration it enables rapid development of industrial applications by reducing the training times. Another critical problem arises with the improper distribution of samples among classes known as the class imbalance problem, which is addressed by employing a simple statistical methodology. Transfer learning is a known technique for training small datasets which transfers pre-trained weights learned on a large dataset. However, during transfer learning, negative transfer learning is a common problem. Therefore, a stepwise transfer learning approach is proposed which can help in fast convergence, while reducing overfitting and preventing negative transfer learning during knowledge transfer across domains. The system is trained and evaluated on two plant disease datasets i.e., PlantVillage (a publicly available dataset) and pepper disease dataset provided by the National Institute of Horticultural and Herbal Science, Republic of Korea. The pepper dataset is particularly challenging as it contains images from different parts of the plant such as the leaf, pulp, and stem. The proposed system has dominated the previous works on the PlantVillage dataset and achieved 99% and 99.69% accuracy on the Pepper dataset and PlantVillage datasets, respectively.

A plethora of second-language (L2) acquisition studies have tested the acquisition of English determiners in relation to simple noun phrases (NPs) of the generic versus non-generic type (e.g., Ionin., et al., 2008; Momenzadea & Youhanaeeb, 2014) rather than complex NPs, such as the ‘of-phrase’ construction, as in ‘the concept of love’. This study, therefore, contributes to our knowledge by addressing how first-language (L1) transfer from Jordanian-Arabic or Cypriot-Greek to English may influence the use of the definite and zero articles in relation to the configuration of the ‘of-phrase’ construction. This construction is comprised of two nominals (Ns): N1: definite noun+of+N2: bare noun. However, the definiteness feature in the equivalent L1 constructions is realised via an Arabic syntactic construct phrase and a Greek spreading feature. The statistical analyses of the data collected by a questionnaire and a forced-choice elicitation task indicated misuse of the before both Ns by the L2 groups at lower proficiency levels of English and provided evidence of L1 negative transfer, which resulted from the syntax-semantics mismatch between the participants’ L1s and L2. The results supported the Feature Reassembly Hypothesis (Lardiere, 2008, 2009). With the increase in the quality and quantity of input in the form of daily exposure to English, the L2 groups figured out how to restructure their L1 features to match the configuration of the L2 regarding their use of the. The age of the Jordanian participants and exposure to English at university/school/work of both groups influenced the use of zero and the; respectively.

Prose, containing colorful Chinese culture, is one of the most important treasures of Chinese literature. And in order to spread it abroad, the translation of it plays an indispensable role, because the good English version helps foreigners to understand more about Chinese culture. However, due to the different environments as well as thinking models between English and Chinese, translators are affected by the Negative Transfer, and couldn’t successfully transform their thinking patterns and thus bring about various kinds of Chinglish while in the practice of Chinese-English translation of prose. The subject of this paper is dozens of prose translations carefully completed by students’ cooperation in groups. On the basis that the translations are basically free of grammatical errors, the author explores Chinglish in those translations from the perspective of Negative Transfer, focusing on the main categories and causes of Chinglish at lexical and syntactic levels, and putting forward relative countermeasures.

Data sparsity is a major challenge for collaborative filtering recommender systems. A promising solution is to utilize feedback or ratings from multiple domains to improve the performance of recommendations in a collective way, known as the cross-domain recommendation. Cross-domain recommendation using heterogeneous feedback is a popular solution, which transfers knowledge from the more easily available auxiliary binary feedback to improve the prediction performance of the target domain. Most of the existing work focuses on the transfer of knowledge between different domains from the same website, where user behavior data in different domains can be fully shared. The existing work mainly assumes that data from different domains can be fully shared. However, due to the constraints of business privacy policies, it is difficult to directly share exactly the same user behavior data between different e-commerce websites. It results in that the user’s latent factors learned in the auxiliary domain cannot be directly transferred to the target domain, otherwise, it will cause a negative transfer issue. In this article, we consider that the auxiliary domain with numerical ratings and target domains with binary feedbacks only share overlapping items rather than users. We propose a Selective Knowledge Transfer for Cross-domain Collaborative Recommendation, called SKT. The proposed SKT framework not only transfers the item’s latent factors learned from the auxiliary domain to the target domain, but also selectively transfers the user’s latent factors learned from the auxiliary domain to the target domain. In addition, due to the introduction of co-graph regularization of user graphs and item graphs, SKT can maintain respective intrinsic geometric structure within each domain and thus avoid negative transfer issue. Extensive experiments conducted on two real-world datasets show that our SKT method is significantly better than all baseline methods at various density levels.

This article is intended to analyze Chinese English learners’ use of “wind” and compare it with the use of “wind” by English speakers. Similarities and differences of conceptual metaphor of the two “winds” are analyzed. The main research method is to use corpus: COCA (Corpus of Contemporary American English) and CLEC (Chinese learners in the Chinese Learner English Corpus). Besides, qualitative method is involved in the analysis and discussion. The research found that there are both similarities and differences in conceptual metaphor of “wind” by English speakers and “wind” by Chinese English learners, due to both positive and negative transfer from “feng” in Chinese to “wind” In English. This research is aimed at giving a better understanding of conceptual metaphor of vocabulary and suggesting some effective strategies in foreign language vocabulary teaching.

After the change of sovereignty from the United Kingdom to People’s Republic of China, Mandarin has become increasingly important in Hong Kong society. However, it is observed that many native Cantonese speakers have problems mastering Mandarin tones despite the less complex tonal system. This study attempts to investigate the Mandarin tonal error patterns made by Hong Kong learners and the possible causes for these errors. A reading aloud test and a follow-up interview were designed to collect the data. The findings of the study show that a large number of students were confused about Mandarin tones and found them difficult to master, especially between tone 1 and tone 4 and tone 2 and tone 3. All the participants made these two mistakes in the test. This study hypothesizes that L1 negative transfer is the key factor that helps Cantonese speaking learners’ ability to establish a new tone in the L2. However, the reason for participants’ tonal mistakes is not due to one single cause; instead, the errors are caused by multiple factors, including lack of tonal knowledge, incomplete tonal acquisition, and tonal errors explained by the Speech Learning Model.

The aim of this study is to find out to what extent Turkish EFL students make interlingual (interference) and intralingual (developmental) errors in writing at the university level. The mid-term and final examination papers of 50 Turkish EFL students were taken and their errors were categorized into interference and developmental errors. The results of our research have shown that they made a mean of %14.6 interference errors. Thus, this proportion does not confirm the contrastive analysis hypothesis claiming that all errors are due to the negative transfer from the mother tongue. On the other hand, they made a mean of %85.4 developmental errors which are not related to L1. So the results of T?Test indicate that Turkish EFL students made significantly more (p

In today’s world, educational contexts are getting increasingly multicultural. Although EFL (English as a Foreign Language) classes in Turkey were mostly composed of Turkish students a decade ago, today students from any part of the world are brought together within the frame of various student exchange programs. In Turkish EFL contexts, students coming from Middle Eastern countries make up a huge part of classes (A–zer, 2016). The objective of the current study is to examine L1 negative interference errors of Turkish and Arabic EFL learners. Accordingly, 30 B1 (Intermediate) EFL students’ written assignments were examined and divided into categories. Besides, this study makes a comparative analysis of Turkish and Arabic EFL learners' grammatical and lexical errors within the frame of L1 negative transfer. When the results of the Mann-Whitney U test are analyzed, it is understood that Turkish and Arabic students' grammatical interference exam scores differ statistically (p

This study was conducted to test the effects of digital games on Asian students’ English vocabulary learning. Data used in this study were collected from Web of Science (WoS) and Education Resources Information Center (ERIC) from 2008 to 2018. A meta-analysis was performed to minimize heterogeneity between studies. The data were analyzed and visualized according to effect sizes. The results of this study showed that digital games played a significant role in promoting Asian students’ English vocabulary learning. This study also examined five factors that influenced students’ English vocabulary learning, including learning stage, game-aided teaching method, native language background, game platform, and game scenario. Effects of these factors were compared by referring to the standards for effect sizes. The result indicated that native language background had the most prominent impact, while negative transfer of native language hindered the vocabulary learning of Asian students. The effects of game platform, game-aided teaching method, and game scenario were not significant. The implication of this study is that creating context-based, high-quality educational games is a promising approach to promote Asian students’ English language learning.

Open Set Domain Adaptation (OSDA) focuses on bridging the domain gap between a labeled source domain and an unlabeled target domain, while also rejecting target classes that are not present in the source as unknown. The challenges of this task are closely related to those of Positive-Unlabeled (PU) learning where it is essential to discriminate between positive (known) and negative (unknown) class samples in the unlabeled target data. With this newly discovered connection, we leverage the theoretical framework of PU learning for OSDA and, at the same time, we extend PU learning to tackle uneven data distributions. Our method combines domain adversarial learning with a new non-negative risk estimator for PU learning based on self-supervised sample reconstruction. With experiments on digit recognition and object classification, we validate our risk estimator and demonstrate that our approach allows reducing the domain gap without suffering from negative transfer.

In this paper we propose two methods for instance transfer based on conformal prediction. As a distinctive character, both of the methods are model independent and combine feature selection and source-instance selection to avoid negative transfer. The methods have been tested experimentally for different types of classification model on several benchmark data sets. The experimental results demonstrate that the new methods are capable of outperforming significantly standard instance transfer methods. (C) 2019 Elsevier B.V. All rights reserved.

The transfer learning methodology leverages knowledge from the source domain with abundant training data to the insufficient target domain. Recently, new approaches continue to be developed and used to solve different classification tasks, ranging from public news to videos and to many others. Most transfer learning methods are based on the assumption that both source and target data are in the same feature space or with the same data distribution, which however is not always true in real applications where it would lead to a negative transfer. In order to overcome this hurdle, the multiple-source transfer learning framework is useful. Since many real systems can be represented by networks, how to utilize the structural similarity between different networks so as to increase the transfer effectiveness becomes important. In this paper, the NES specification index is used to quantitatively measure the structural similarity between two networks, based on which a new transfer learning method (named NES-TL) is developed. Experiments on tag popularity prediction in StackExchange Q&A communities verify the effectiveness of the proposed approach, showing that it behaves better than existing baseline methods.

A category within lexical transfer, collocation transfer is a significant phenomenon within language production which may have negative effects on writing quality when demonstrated in the form of negative transfer. However, studies on collocation transfer in the texts written by writers of English as a Foreign Language appear to be somewhat limited. In this respect, this study aimed to reveal if the frequently used lexical collocations by undergraduate learners of English with a Turkish L1 background in their written production demonstrated negative collocation transfer. To this end, a corpus of 160 literary analysis essays written by 2nd year undergraduate students of an English Language Teaching (ELT) department in Turkey were investigated to trace negative collocation transfer from Turkish to English. The Detection, Explanation, Evaluation model was used for the analysis of the collocations. The findings indicated that around a quarter of the collocations used by the writers in their texts were negatively transferred from Turkish. The results highlighted the importance of teaching collocations in EFL contexts.

L’article examine les transferts positifs et négatifs entre le tchèque et le français, en se focalisant sur les prépositions françaises sur et sous et leurs équivalents tchèques. D’abord, nous décrivons les constructions étudiées et émettons des hypothèses concernant leur acquisition. Puis, nous présentons les résultats d’une enquête linguistique, contenant une tâche expérimentale ciblant les deux prépositions, qui a été effectuée auprès des étudiants tchèques du français au printemps 2019. Ensuite, nous mettons en lumière certaines données intéressantes et déterminons les transferts positifs et négatifs entre le français et le tchèque au sujet des prépositions sur et sous.

Person re-identification (Re-ID) across multiple datasets is a challenging task due to two main reasons: the presence of large cross-dataset distinctions and the absence of annotated target instances. To address these two issues, this paper proposes a domain adaptive attention learning approach to reliably transfer discriminative representation from the labeled source domain to the unlabeled target domain. In this approach, a domain adaptive attention model is learned to separate the feature map into domain-shared part and domain-specific part. In this manner, the domain-shared part is used to capture transferable cues that can compensate cross-dataset distinctions and give positive contributions to the target task, while the domain-specific part aims to model the noisy information to avoid the negative transfer caused by domain diversity. A soft label loss is further employed to take full use of unlabeled target data by estimating pseudo labels. Extensive experiments on the Market-1501, DukeMTMC-reID and MSMT17 benchmarks demonstrate the proposed approach outperforms the state-of-the-arts.

Learner’s first language (L1) could influence the learning process of a second (L2) language across the different linguistic levels. The pronunciation of second languages usually accented by the first language phonetic system. This accented speech impacts learners’ expressive, and receptive skills to different extents (Derwing and Munro 2005)[i]. Many researchers mentioned that the evaluation of the process of learning L2 shouldn’t rely neither on the amount of the learned vocabulary nor on the grammar. Otherwise, they focused on the importance of acquiring an accurate pronunciation for a mutual intelligible communication between non-natives and naitive speakers. (Derwing & Munro, 2015;)[ii] (Levis, 2018 [iii]). Accurate pronunciation of L2 governs by many factors such as the extent of the linguistic difference and similarity between L1 and L2. This extent could subsequently influence the type and degree of L1 transfer into L2 pronunciation. Transfer, could be either negative or positive. In the case of lacking the L1 phonetic and phonological system to some features of L2, the interference of L1 into L2 could be regarded as negative transfer. These dissimilarities my cause perception and production challenges to the L2 learners. [i] Derwing and Munro 2005 Accent and Pronunciation Teaching: A Research Based Approach TESOL Quarterly Journal Volume39, Issue3Pages 379-397 [ii] TRACEY M.Derwing & MURRAY J. MUNRO (2015) pronunciation Fundamentals: Evidence-based Perspectives for L2 Teaching and Research (Language Learning & Language Teaching) John Benjamins Publishing Company; UK ed. Edition. [iii] John M. Levis (2018), intelligibility, Oral communication, and the teaching of Pronunciation, Cambridge University Press

Language transfer means the application of knowledge and understanding of a source language on a target language. Dissimilarities between the source language and the target language could result in a negative transfer. This essay aims to investigate the interference of Indonesian language on English second language writing regarding the order of noun phrase constituents, number marking, and verb tenses. The results presented in this essay were based on three elicitation tasks that were answered by 20 Indonesian students who studied English at a senior high school in Sumbawa Regency, Indonesia. Their ages were between 15 to 18 years old. The participants were given a test that consisted of three elicitation tasks focusing on sentence writing. The results showed that the negative transfer was still found in the students’ writing. However, the students may make mistakes since they had a lack of knowledge, or because they were unable to apply what they knew in their writing.

Transfer learning approach can be applied to electric load forecasting because electric load data from nearby locations are significantly correlated. However, ordinary transfer learning methods may bring negative transfer into load forecasting as time series prediction is not exactly the same as traditional data regression problem. Consequently, this paper proposes a novel hybrid transfer learning model based on time series decomposition. Firstly, trend and seasonal components are handled by standard machine learning approach so that seasonal cycles of electric load data can be interpreted better. Secondly, two-stage transfer regression model is established to forecast the irregular component in order to improve the forecasting accuracy. The negative transfer is successfully avoided, and the prediction accuracies are significantly improved because of time series decomposing and the additional information provided by the related dataset. The case study presented by two real-world power load datasets illustrates that the proposed approach can improve electric load prediction for a location by 30% at most by using additional data from another location.