Transfer Learning Tasks Research Articles

Investigating Transfer Learning in Noisy Environments: A Study of Predecessor and Successor Features in Spatial Learning Using a T-Maze.

In this study, we investigate the adaptability of artificial agents within a noisy T-maze that use Markov decision processes (MDPs) and successor feature (SF) and predecessor feature (PF) learning algorithms. Our focus is on quantifying how varying the hyperparameters, specifically the reward learning rate (αr) and the eligibility trace decay rate (λ), can enhance their adaptability. Adaptation is evaluated by analyzing the hyperparameters of cumulative reward, step length, adaptation rate, and adaptation step length and the relationships between them using Spearman's correlation tests and linear regression. Our findings reveal that an αr of 0.9 consistently yields superior adaptation across all metrics at a noise level of 0.05. However, the optimal setting for λ varies by metric and context. In discussing these results, we emphasize the critical role of hyperparameter optimization in refining the performance and transfer learning efficacy of learning algorithms. This research advances our understanding of the functionality of PF and SF algorithms, particularly in navigating the inherent uncertainty of transfer learning tasks. By offering insights into the optimal hyperparameter configurations, this study contributes to the development of more adaptive and robust learning algorithms, paving the way for future explorations in artificial intelligence and neuroscience.

FireDA: A Domain Adaptation-Based Method for Forest Fire Recognition with Limited Labeled Scenarios

Vision-based forest fire detection systems have significantly advanced through Deep Learning (DL) applications. However, DL-based models typically require large-scale labeled datasets for effective training, where the quality of data annotation is crucial to their performance. To address challenges related to the quality and quantity of labeling, a domain adaptation-based approach called FireDA is proposed for forest fire recognition in scenarios with limited labels. Domain adaptation, a subfield of transfer learning, facilitates the transfer of knowledge from a labeled source domain to an unlabeled target domain. The construction of the source domain FBD is initiated, which includes three common fire scenarios: forest (F), brightness (B), and darkness (D), utilizing publicly available labeled data. Subsequently, a novel algorithm called Neighborhood Aggregation-based 2-Stage Domain Adaptation (NA2SDA) is proposed. This method integrates feature distribution alignment with target domain Proxy Classification Loss (PCL), leveraging a neighborhood aggregation mechanism and a memory bank designed for the unlabeled samples in the target domain. This mechanism calibrates the source classifier and generates more accurate pseudo-labels for the unlabeled sample. Consequently, based on these pseudo-labels, the Local Maximum Mean Discrepancy (LMMD) and the Proxy Classification Loss (PCL) are computed. To validate the efficacy of the proposed method, the publicly available forest fire dataset, FLAME, is employed as the target domain for constructing a transfer learning task. The results demonstrate that our method achieves performance comparable to the supervised Convolutional Neural Network (CNN)-based state-of-the-art (SOTA) method, without requiring access to labels from the FLAME training set. Therefore, our study presents a viable solution for forest fire recognition in scenarios with limited labeling and establishes a high-accuracy benchmark for future research.

Simultaneously predicting SPAD and water content in rice leaves using hyperspectral imaging with deep multi-task regression and transfer component analysis.

Water content and chlorophyll content are important indicators for monitoring rice growth status. Simultaneous detection of water content and chlorophyll content is of significance. Different varieties of rice show differences in phenotype, resulting in the difficulties of establishing a universal model. In this study, hyperspectral imaging was used to detect the Soil and Plant Analyzer Development (SPAD) values and water content of fresh rice leaves of three rice varieties (Jiahua 1, Xiushui 121 and Xiushui 134). Both partial least squares regression and convolutional neural networks were used to establish single-task and multi-task models. Transfer component analysis (TCA) was used as transfer learning to learn the common features to achieve an approximate identical distribution between any two varieties. Single-task and multi-task models were also built using the features of the source domain, and these models were applied to the target domain. These results indicated that for models of each rice variety the prediction accuracy of most multi-task models was close to that of single-task models. As for TCA, the results showed that the single-task model achieved good performance for all transfer learning tasks. Compared with the original model, good and differentiated results were obtained for the models using features learned by TCA for both the source domain and target domain. The multi-task models could be constructed to predict SPAD values and water content simultaneously and then transferred to another rice variety, which could improve the efficiency of model construction and realize rapid detection of rice growth indicators. © 2024 Society of Chemical Industry.

Active Learning for Rapid Targeted Synthesis of Compositionally Complex Alloys.

The next generation of advanced materials is tending toward increasingly complex compositions. Synthesizing precise composition is time-consuming and becomes exponentially demanding with increasing compositional complexity. An experienced human operator does significantly better than a novice but still struggles to consistently achieve precision when synthesis parameters are coupled. The time to optimize synthesis becomes a barrier to exploring scientifically and technologically exciting compositionally complex materials. This investigation demonstrates an active learning (AL) approach for optimizing physical vapor deposition synthesis of thin-film alloys with up to five principal elements. We compared AL-based on Gaussian process (GP) and random forest (RF) models. The best performing models were able to discover synthesis parameters for a target quinary alloy in 14 iterations. We also demonstrate the capability of these models to be used in transfer learning tasks. RF and GP models trained on lower dimensional systems (i.e., ternary, quarternary) show an immediate improvement in prediction accuracy compared to models trained only on quinary samples. Furthermore, samples that only share a few elements in common with the target composition can be used for model pre-training. We believe that such AL approaches can be widely adapted to significantly accelerate the exploration of compositionally complex materials.

Fault Diagnosis Method of Special Vehicle Bearing Based on Multi-Scale Feature Fusion and Transfer Adversarial Learning.

To address the issues of inadequate feature extraction for rolling bearings, inaccurate fault diagnosis, and overfitting in complex operating conditions, this paper proposes a rolling bearing diagnosis method based on multi-scale feature fusion and transfer adversarial learning. Firstly, a multi-scale convolutional fusion layer is designed to effectively extract fault features from the original vibration signals at multiple time scales. Through a feature encoding fusion module based on the multi-head attention mechanism, feature fusion extraction is performed, which can model long-distance contextual information and significantly improve diagnostic accuracy and anti-noise capability. Secondly, based on the domain adaptation (DA) cross-domain feature adversarial learning strategy of transfer learning methods, the extraction of optimal domain-invariant features is achieved by reducing the gap in data distribution between the target domain and the source domain, addressing the call for research on fault diagnosis across operating conditions, equipment, and virtual-real migrations. Finally, experiments were conducted to verify and optimize the effectiveness of the feature extraction and fusion network. A public bearing dataset was used as the source domain data, and special vehicle bearing data were selected as the target domain data for comparative experiments on the effect of network transfer learning. The experimental results demonstrate that the proposed method exhibits an exceptional performance in cross-domain and variable load environments. In multiple bearing cross-domain transfer learning tasks, the method achieves an average migration fault diagnosis accuracy rate of up to 98.65%. When compared with existing methods, the proposed method significantly enhances the ability of data feature extraction, thereby achieving a more robust diagnostic performance.

Hybrid modeling for improved extrapolation and transfer learning in the chemical processing industry

In the Chemical Processing Industry, shifts in market demand often require the implementation of new operational modes that balance economic advantages with the need to meet quality and sustainability targets. Hybrid modeling can support these process changes by combining physics-based and data-driven modeling principles to improve prediction accuracy and interpretation. In this work, the use of hybrid modeling for improved extrapolation and transfer learning activities is examined in the context of simulated biodiesel production. We study and compare different configurations of hybrid modeling, including the parallel and serial structures. We also compare hybrid modeling approaches with physics-based and data-driven models, and find that hybrid modeling consistently outperforms these benchmarks in both extrapolation and transfer learning tasks. Hybrid modeling also requires fewer samples than other benchmarks for the transfer learning task. These results suggest that hybrid modeling is an effective approach for supporting decision-making and optimizing process changes in the CPI.

Selective Random Walk for Transfer Learning in Heterogeneous Label Spaces.

Transfer learning has been widely used in different scenarios, especially in those lacking enough labeled data. However, most of the existing transfer learning methods are based on the assumption that the source and target domains should share the label space entirely or partially, which greatly limits their application scopes. In this article, a Selective Random Walk (SRW) method for transfer learning in heterogeneous label spaces is proposed to make full use of unlabeled auxiliary data, which acts as a bridge for knowledge transfer from the source domain to the target domain. The proposed SRW method can explicitly identify transfer sequences between source and target instances via auxiliary instances based on random walk techniques. Since not all of the transfer sequences generated by random walk are credible for the target task, the SRW method can learn to weight transfer sequences adaptively. Based on the weights of the transfer sequences, the SRW method leverages knowledge by forcing adjacent data points in the transfer sequence to be similar and making the target data point in the sequence represented by other data points in the same sequence. Experiments show that the SRW method outperforms state-of-the-art models in plenty of transfer learning tasks with heterogeneous label spaces constructed within and across several benchmark datasets.

Advancing language models through domain knowledge integration: a comprehensive approach to training, evaluation, and optimization of social scientific neural word embeddings

This article proposes a comprehensive strategy for training, evaluating, and optimizing domain-specific word2vec-based word embeddings, using social science literature as an example. Our primary objectives are: (1) to train the embeddings utilizing a corpus of social science text, (2) to test their performance against domain-unspecific embeddings using our developed intrinsic and extrinsic evaluation strategy, and (3) to enhance their performance even further by using domain knowledge. As an integral part of this approach, we present SociRel-461, a domain-knowledge dictionary designed for the intrinsic evaluation and subsequent refinement of social science word embeddings. Using a dataset of 100,000 full-text scientific articles in sociology, we train multiple vector space models, which we then benchmark against a larger, pre-trained general language embedding model as part of our extrinsic evaluation. Furthermore, we developed a transfer learning multi-label classification task for extrinsic evaluation. Our findings reveal that domain-specific embeddings outperform their domain-unspecific counterparts in both intrinsic and extrinsic evaluations. We also investigated the retrofitting post-processing method to enhance domain-unspecific embeddings with the domain knowledge embedded in SociRel-461. While retrofitting does not enhance our domain-specific vector space models, it significantly improves the performance of the domain-unspecific embeddings. This highlights the potential of retrofitting for the transfer of domain knowledge to domain-unspecific embeddings. Our results emphasize the importance of utilizing domain-specific word embeddings for better performance in domain specific transfer learning tasks, as they outperform conventional embeddings trained on everyday language.

Research on sound quality of roller chain transmission system based on multi-source transfer learning

To establish the sound quality evaluation model of roller chain transmission system, we collect the running noise under different working conditions. After the noise samples are preprocessed, a group of experienced testers are organized to evaluate them subjectively. Mel frequency cepstral coefficient (MFCC) of each noise sample is calculated, and the MFCC feature map is used as an objective evaluation. Combining with the subjective and objective evaluation results of the roller chain system noise, we can get the original dataset of its sound quality research. However, the number of high-quality noise samples is relatively small. Based on the sound quality research of various chain transmission systems, a novel method called multi-source transfer learning convolutional neural network (MSTL-CNN) is proposed. By transferring knowledge from multiple source tasks to target task, the difficulty of small sample sound quality prediction is solved. Compared with the problem that single source task transfer learning has too much error on some samples, MSTL-CNN can give full play to the advantages of all transfer learning models. The results also show that the MSTL-CNN proposed in this paper is significantly better than the traditional sound quality evaluation methods.

Fast and effective molecular property prediction with transferability map.

Effective transfer learning for molecular property prediction has shown considerable strength in addressing insufficient labeled molecules. Many existing methods either disregard the quantitative relationship between source and target properties, risking negative transfer, or require intensive training on target tasks. To quantify transferability concerning task-relatedness, we propose Principal Gradient-based Measurement (PGM) for transferring molecular property prediction ability. First, we design an optimization-free scheme to calculate a principal gradient for approximating the direction of model optimization on a molecular property prediction dataset. We have analyzed the close connection between the principal gradient and model optimization through mathematical proof. PGM measures the transferability as the distance between the principal gradient obtained from the source dataset and that derived from the target dataset. Then, we perform PGM on various molecular property prediction datasets to build a quantitative transferability map for source dataset selection. Finally, we evaluate PGM on multiple combinations of transfer learning tasks across 12 benchmark molecular property prediction datasets and demonstrate that it can serve as fast and effective guidance to improve the performance of a target task. This work contributes to more efficient discovery of drugs, materials, and catalysts by offering a task-relatedness quantification prior to transfer learning and understanding the relationship between chemical properties.

FTSDC: A novel federated transfer learning strategy for bearing cross-machine fault diagnosis based on dual-correction training

In recent years, although traditional intelligent fault diagnosis methods have achieved satisfactory development in transfer learning tasks, the sample information that the single client can generally provide is extremely limited in real industrial scenarios. And the private data needs to be guaranteed not to leave the local storage during the application process, which leads to obstacles for fault diagnosis methods to solve cross-device and cross-scenario client transfer tasks. Therefore, a novel federated transfer learning strategy based on dual-correction training (FTSDC) is proposed, which enables fault diagnosis for multi-device tasks without target domain samples to participate in model training. The dual correction training of source client is proposed to enhance the local network’s generality, which involves two links: The multiple-functions correction model hyperparameters link and the transition training correction network attention area link. Multiple sets of optimization functions are introduced into local network training to reduce the covariate drift phenomenon caused by domain discrepancies. And the local model fine-tunes the parameters of the trained network through the transition dataset to correct attention areas. Furthermore, The central server evaluates each source client model according to the contribution of the local model to the transfer strategy, and the local model parameters are cumulatively optimized by referring to the diagnosic experience of the global model. It is worth noting that the trained local model can still demonstrate commendable performance even when faced with more complex scenarios. Finally, the proposed scheme was rigorously assessed against diverse federated learning methods across various task settings during the experimental session. The results confirmed the effectiveness of the proposed scheme in safeguarding the data privacy of each client while concurrently enhancing the diagnostic accuracy of the target task.

Genetic Programming for Document Classification: A Transductive Transfer Learning System.

Document classification is a challenging task to the data being high-dimensional and sparse. Many transfer learning methods have been investigated for improving the classification performance by effectively transferring knowledge from a source domain to a target domain, which is similar to but different from the source domain. However, most of the existing methods cannot handle the case that the training data of the target domain does not have labels. In this study, we propose a transductive transfer learning system, utilizing solutions evolved by genetic programming (GP) on a source domain to automatically pseudolabel the training data in the target domain in order to train classifiers. Different from many other transfer learning techniques, the proposed system pseudolabels target-domain training data to retrains classifiers using all target-domain features. The proposed method is examined on nine transfer learning tasks, and the results show that the proposed transductive GP system has better prediction accuracy on the test data in the target domain than existing transfer learning approaches including subspace alignment-domain adaptation methods, feature-level-domain adaptation methods, and one latest pseudolabeling strategy-based method.

CLEAR: Cluster-Enhanced Contrast for Self-Supervised Graph Representation Learning.

This article studies self-supervised graph representation learning, which is critical to various tasks, such as protein property prediction. Existing methods typically aggregate representations of each individual node as graph representations, but fail to comprehensively explore local substructures (i.e., motifs and subgraphs), which also play important roles in many graph mining tasks. In this article, we propose a self-supervised graph representation learning framework named cluster-enhanced Contrast (CLEAR) that models the structural semantics of a graph from graph-level and substructure-level granularities, i.e., global semantics and local semantics, respectively. Specifically, we use graph-level augmentation strategies followed by a graph neural network-based encoder to explore global semantics. As for local semantics, we first use graph clustering techniques to partition each whole graph into several subgraphs while preserving as much semantic information as possible. We further employ a self-attention interaction module to aggregate the semantics of all subgraphs into a local-view graph representation. Moreover, we integrate both global semantics and local semantics into a multiview graph contrastive learning framework, enhancing the semantic-discriminative ability of graph representations. Extensive experiments on various real-world benchmarks demonstrate the efficacy of the proposed over current graph self-supervised representation learning approaches on both graph classification and transfer learning tasks.

Enhancing Intracranial Hemorrhage Diagnosis through Deep Learning Models

Timely diagnosis is crucial for the successful treatment of a serious medical condition like brain hemorrhage. Deep learning algorithms have shown great promise in applications for medical image analysis, like the identification of brain hemorrhages. The goal of this study is to assess how well various deep learning algorithms can detect brain hemorrhages. Using a suitable dataset, the study evaluates the computational efficiency, accuracy, sensitivity, and specificity of the selected algorithms. The results demonstrate the potential of deep learning models to assist physicians in identifying this potentially fatal condition and demonstrate how well they can identify brain hemorrhages. The study’s findings improve automated brain hemorrhage detection technology, improving patient outcomes and the efficiency of healthcare delivery. EfficientNetB3 typically achieves higher accuracy due to its increased model complexity. Despite its increased complexity, EfficientNetB3 is still more parameter-efficient and computationally efficient than many alternative architectures. EfficientNetB3’s strong performance and feature extraction capabilities make it a good choice for transfer learning tasks. Out of all the models implemented in this paper, the proposed model with EfficientNetB3 gave the best accuracy for training as well as validation i.e. 99.95% and 93.29% respectively followed by EfficientNetB2, ResNet, SEResNext and ResNext.

Transfer learning for a foundational chemistry model.

Data-driven chemistry has garnered much interest concurrent with improvements in hardware and the development of new machine learning models. However, obtaining sufficiently large, accurate datasets of a desired chemical outcome for data-driven chemistry remains a challenge. The community has made significant efforts to democratize and curate available information for more facile machine learning applications, but the limiting factor is usually the laborious nature of generating large-scale data. Transfer learning has been noted in certain applications to alleviate some of the data burden, but this protocol is typically carried out on a case-by-case basis, with the transfer learning task expertly chosen to fit the finetuning. Herein, I develop a machine learning framework capable of accurate chemistry-relevant prediction amid general sources of low data. First, a chemical "foundational model" is trained using a dataset of ∼1 million experimental organic crystal structures. A task specific module is then stacked atop this foundational model and subjected to finetuning. This approach achieves state-of-the-art performance on a diverse set of tasks: toxicity prediction, yield prediction, and odor prediction.

Transferability-Guided Cross-Domain Cross-Task Transfer Learning.

We propose two novel transferability metrics fast optimal transport-based conditional entropy (F-OTCE) and joint correspondence OTCE (JC-OTCE) to evaluate how much the source model (task) can benefit the learning of the target task and to learn more generalizable representations for cross-domain cross-task transfer learning. Unlike the original OTCE metric that requires evaluating the empirical transferability on auxiliary tasks, our metrics are auxiliary-free such that they can be computed much more efficiently. Specifically, F-OTCE estimates transferability by first solving an optimal transport (OT) problem between source and target distributions and then uses the optimal coupling to compute the negative conditional entropy (NCE) between the source and target labels. It can also serve as an objective function to enhance downstream transfer learning tasks including model finetuning and domain generalization (DG). Meanwhile, JC-OTCE improves the transferability accuracy of F-OTCE by including label distances in the OT problem, though it incurs additional computation costs. Extensive experiments demonstrate that F-OTCE and JC-OTCE outperform state-of-the-art auxiliary-free metrics by 21.1% and 25.8% , respectively, in correlation coefficient with the ground-truth transfer accuracy. By eliminating the training cost of auxiliary tasks, the two metrics reduce the total computation time of the previous method from 43 min to 9.32 and 10.78 s, respectively, for a pair of tasks. When applied in the model finetuning and DG tasks, F-OTCE shows significant improvements in the transfer accuracy in few-shot classification experiments, with up to 4.41% and 2.34% accuracy gains, respectively.

Self-supervised Graph-level Representation Learning with Adversarial Contrastive Learning

The recently developed unsupervised graph representation learning approaches apply contrastive learning into graph-structured data and achieve promising performance. However, these methods mainly focus on graph augmentation for positive samples, while the negative mining strategies for graph contrastive learning are less explored, leading to sub-optimal performance. To tackle this issue, we propose a Graph Adversarial Contrastive Learning (GraphACL) scheme that learns a bank of negative samples for effective self-supervised whole-graph representation learning. Our GraphACL consists of (i) a graph encoding branch that generates the representations of positive samples and (ii) an adversarial generation branch that produces a bank of negative samples. To generate more powerful hard negative samples, our method minimizes the contrastive loss during encoding updating while maximizing the contrastive loss adversarially over the negative samples for providing the challenging contrastive task. Moreover, the quality of representations produced by the adversarial generation branch is enhanced through the regularization of carefully designed bank divergence loss and bank orthogonality loss. We optimize the parameters of the graph encoding branch and adversarial generation branch alternately. Extensive experiments on 14 real-world benchmarks on both graph classification and transfer learning tasks demonstrate the effectiveness of the proposed approach over existing graph self-supervised representation learning methods.

Transfer Learning With Singular Value Decomposition of Multichannel Convolution Matrices.

The task of transfer learning using pretrained convolutional neural networks is considered. We propose a convolution-SVD layer to analyze the convolution operators with a singular value decomposition computed in the Fourier domain. Singular vectors extracted from the source domain are transferred to the target domain, whereas the singular values are fine-tuned with a target data set. In this way, dimension reduction is achieved to avoid overfitting, while some flexibility to fine-tune the convolution kernels is maintained. We extend an existing convolution kernel reconstruction algorithm to allow for a reconstruction from an arbitrary set of learned singular values. A generalization bound for a single convolution-SVD layer is devised to show the consistency between training and testing errors. We further introduce a notion of transfer learning gap. We prove that the testing error for a single convolution-SVD layer is bounded in terms of the gap, which motivates us to develop a regularization model with the gap as the regularizer. Numerical experiments are conducted to demonstrate the superiority of the proposed model in solving classification problems and the influence of various parameters. In particular, the regularization is shown to yield a significantly higher prediction accuracy.

Transferring Vision-Language Models for Visual Recognition: A Classifier Perspective

Transferring knowledge from pre-trained deep models for downstream tasks, particularly with limited labeled samples, is a fundamental problem in computer vision research. Recent advances in large-scale, task-agnostic vision-language pre-trained models, which are learned with billions of samples, have shed new light on this problem. In this study, we investigate how to efficiently transfer aligned visual and textual knowledge for downstream visual recognition tasks. We first revisit the role of the linear classifier in the vanilla transfer learning framework, and then propose a new paradigm where the parameters of the classifier are initialized with semantic targets from the textual encoder and remain fixed during optimization. To provide a comparison, we also initialize the classifier with knowledge from various resources. In the empirical study, we demonstrate that our paradigm improves the performance and training speed of transfer learning tasks. With only minor modifications, our approach proves effective across 17 visual datasets that span three different data domains: image, video, and 3D point cloud.

Transferable driver facial expression recognition based on joint discriminative correlation alignment network with enhanced feature attention

AbstractFacial expression is closely related to the emotions of drivers, thus facilitating safe driving detection in advanced driving assistance system (ADAS). Recently, deep learning techniques have become prevalent for facial expression recognition. However, for driving scenarios, the facial expression recognition is mainly challenged by the problems of small sample size as well as effective feature representation. To address the above issues, this paper puts forward a transfer learning‐based method for driver facial expression recognition by fully exploiting the other sources of facial expression data that may obey different distributions. Specifically, an enhanced feature attention module is firstly devised such that affluent features with multi‐scales can be extracted and refined based on the spatial and channel attention mechanisms. Then, a joint correlation alignment loss is presented ensuring that the samples in the source and target domains are transformed into the shared common subspace to reduce difference of both the marginal and conditional distributions. Multiple transfer learning tasks on real‐world data are carried out to evaluate the proposed method. The experimental results show that our model achieves better recognition accuracy for driver facial expression compared with several traditional and deep learning‐based transfer learning algorithms.