An official website of the United States government
There is an increasing number of pre-trained deep neural network models. However, it is still unclear how to effectively use these models for a new task. Transfer learning, which aims to transfer knowledge from source tasks to a target task, is an effective solution to this problem. Fine-tuning is a popular transfer learning technique for deep neural networks where a few rounds of training are applied to the parameters of a pre-trained model to adapt them to a new task. Despite its popularity, in this paper we show that fine-tuning suffers from several drawbacks. We propose an adaptive fine-tuning approach, called AdaFilter, which selects only a part of the convolutional filters in the pre-trained model to optimize on a per-example basis. We use a recurrent gated network to selectively fine-tune convolutional filters based on the activations of the previous layer. We experiment with 7 public image classification datasets and the results show that AdaFilter can reduce the average classification error of the standard fine-tuning by 2.54%.
more »
« less
Learning New Tricks From Old Dogs: Multi-Source Transfer Learning From Pre-Trained Networks
The advent of deep learning algorithms for mobile devices and sensors has led to a dramatic expansion in the availability and number of systems trained on a wide range of machine learning tasks, creating a host of opportunities and challenges in the realm of transfer learning. Currently, most transfer learning methods require some kind of control over the systems learned, either by enforcing constraints dur- ing the source training, or through the use of a joint optimization objective between tasks that requires all data be co-located for training. However, for practical, pri- vacy, or other reasons, in a variety of applications we may have no control over the individual source task training, nor access to source training samples. Instead we only have access to features pre-trained on such data as the output of “black-boxes.” For such scenarios, we consider the multi-source learning problem of training a classifier using an ensemble of pre-trained neural networks for a set of classes that have not been observed by any of the source networks, and for which we have very few training samples. We show that by using these distributed networks as feature extractors, we can train an effective classifier in a computationally-efficient manner using tools from (nonlinear) maximal correlation analysis. In particular, we develop a method we refer to as maximal correlation weighting (MCW) to build the required target classifier from an appropriate weighting of the feature functions from the source networks. We illustrate the effectiveness of the resulting classi- fier on datasets derived from the CIFAR-100, Stanford Dogs, and Tiny ImageNet datasets, and, in addition, use the methodology to characterize the relative value of different source tasks in learning a target task.
more »
« less
- Award ID(s):
- 1816209
- PAR ID:
- 10201175
- Date Published:
- Journal Name:
- Advances in neural information processing systems
- ISSN:
- 1049-5258
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Fine-tuning a pre-trained model on a downstream task often degrades its original capabilities, a phenomenon known as "catastrophic forgetting". This is especially an issue when one does not have access to the data and recipe used to develop the pre-trained model. Under this constraint, most existing methods for mitigating forgetting are inapplicable. To address this challenge, we propose a sample weighting scheme for the fine-tuning data solely based on the pre-trained model's losses. Specifically, we upweight the easy samples on which the pre-trained model's loss is low and vice versa to limit the drift from the pre-trained model. Our approach is orthogonal and yet complementary to existing methods; while such methods mostly operate on parameter or gradient space, we concentrate on the sample space. We theoretically analyze the impact of fine-tuning with our method in a linear setting, showing that it stalls learning in a certain subspace which inhibits overfitting to the target task. We empirically demonstrate the efficacy of our method on both language and vision tasks. As an example, when fine-tuning Gemma 2 2B on MetaMathQA, our method results in only a 0.8% drop in accuracy on GSM8K (another math dataset) compared to standard fine-tuning, while preserving 5.4% more accuracy on the pre-training datasets.more » « less
-
null (Ed.)Convolutional Neural Network (CNN) models and many accessible large-scale public visual datasets have brought lots of research work to a remarkable new stage. Benefited from well-trained CNN models, small training datasets can learn comprehensive features by utilizing the preliminary features from transfer learning. However, the performance is not guaranteed when taking these features to construct a new model, as the differences always exist between the source and target domains. In this paper, we propose to build an Evolution Programming-based framework to address various challenges. This framework automates both the feature learning and model building processes. It first identifies the most valuable features from pre-trained models and then constructs a suitable model to understand the characteristic features for different tasks. Each model differs in numerous ways. Overall, the experimental results effectively reach optimal solutions, demonstrating that a time-consuming task could also be conducted by an automated process that exceeds the human ability.more » « less
-
null (Ed.)Abstract: Deep Learning (DL) has made significant changes to a large number of research areas in recent decades. For example, several astonishing Convolutional Neural Network (CNN) models have been built by researchers to fulfill image classification needs using large-scale visual datasets successfully. Transfer Learning (TL) makes use of those pre-trained models to ease the feature learning process for other target domains that contain a smaller amount of training data. Currently, there are numerous ways to utilize features generated by transfer learning. Pre-trained CNN models prepare mid-/high-level features to work for different targeting problem domains. In this paper, a DL feature and model selection framework based on evolutionary programming is proposed to solve the challenges in visual data classification. It automates the process of discovering and obtaining the most representative features generated by the pre-trained DL models for different classification tasks.more » « less
-
null (Ed.)Given its demonstrated ability in analyzing and revealing patterns underlying data, Deep Learning (DL) has been increasingly investigated to complement physics-based models in various aspects of smart manufacturing, such as machine condition monitoring and fault diagnosis, complex manufacturing process modeling, and quality inspection. However, successful implementation of DL techniques relies greatly on the amount, variety, and veracity of data for robust network training. Also, the distributions of data used for network training and application should be identical to avoid the internal covariance shift problem that reduces the network performance applicability. As a promising solution to address these challenges, Transfer Learning (TL) enables DL networks trained on a source domain and task to be applied to a separate target domain and task. This paper presents a domain adversarial TL approach, based upon the concepts of generative adversarial networks. In this method, the optimizer seeks to minimize the loss (i.e., regression or classification accuracy) across the labeled training examples from the source domain while maximizing the loss of the domain classifier across the source and target data sets (i.e., maximizing the similarity of source and target features). The developed domain adversarial TL method has been implemented on a 1-D CNN backbone network and evaluated for prediction of tool wear propagation, using NASA's milling dataset. Performance has been compared to other TL techniques, and the results indicate that domain adversarial TL can successfully allow DL models trained on certain scenarios to be applied to new target tasks.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
