Continual learning with task specialist

Continual learning (CL) adapt the deep learning scenarios with timely updated datasets. However, existing CL models suffer from the catastrophic forgetting issue, where new knowledge replaces past learning. In this paper, we propose Continual Learning with Task Specialists (CLTS) to address the issues of catastrophic forgetting and limited labelled data in real-world datasets by performing class incremental learning of the incoming stream of data. The model consists of Task Specialists (T S) and Task Predictor (T P ) with pre-trained Stable Diffusion (SD) module. Here, we introduce a new specialist to handle a new task sequence and each T S has three blocks; i) a variational autoencoder (V AE) to learn the task distribution in a low dimensional latent space, ii) a K-Means block to perform data clustering and iii) Bootstrapping Language-Image Pre-training (BLIP ) model to generate a small batch of captions from the input data. These captions are fed as input to the pre-trained stable diffusion model (SD) for the generation of task samples. The proposed model does not store any task samples for replay, instead uses generated samples from SD to train the T P module. A comparison study with four SOTA models conducted on three real-world datasets shows that the proposed model outperforms all the selected baselines

XMOL: Explainable Multi-property Optimization of Molecules

Molecular optimization is a key challenge in drug discovery and material science domain, involving the design of molecules with desired properties. Existing methods focus predominantly on single-property optimization, necessitating repetitive runs to target multiple properties, which is inefficient and computationally expensive. Moreover, these methods often lack transparency, making it difficult for researchers to understand and control the optimization process. To address these issues, we propose a novel framework, Explainable Multi-property Optimization of Molecules (XMOL), to optimize multiple molecular properties simultaneously while incorporating explainability. Our approach builds on state-of-the-art geometric diffusion models, extending them to multi-property optimization through the introduction of spectral normalization and enhanced molecular constraints for stabilized training. Additionally, we integrate interpretive and explainable techniques throughout the optimization process. We evaluated XMOL on the real-world molecular datasets i.e., QM9, demonstrating its effectiveness in both single property and multiple properties optimization while offering interpretable results, paving the way for more efficient and reliable molecular design.

U-TELL: Unsupervised Task Expert Lifelong Learning

Continual learning (CL) models are designed to learn new tasks arriving sequentially without re-training the network. However, real-world ML applications have very limited label information and these models suffer from catastrophic forgetting. To address these issues, we propose an unsupervised CL model with task experts called Unsupervised Task Expert Lifelong Learning (U-TELL) to continually learn the data arriving in a sequence addressing catastrophic forgetting. During training of U-TELL, we introduce a new expert on arrival of a new task. Our proposed architecture has task experts, a structured data generator and a task assigner. Each task expert is composed of 3 blocks; i) a variational autoencoder to capture the task distribution and perform data abstraction, ii) a k-means clustering module, and iii) a structure extractor to preserve latent task data signature. During testing, task assigner selects a suitable expert to perform clustering. U-TELL does not store or replay task samples, instead, we use generated structured samples to train the task assigner. We compared U-TELL with five SOTA unsupervised CL methods. U-TELL outperformed all baselines on seven benchmarks and one industry dataset for various CL scenarios with a training time over 6 times faster than the best performing baseline.

Cross-Problem Learning for Solving Vehicle Routing Problems

Existing neural heuristics often train a deep architecture from scratch for each specific vehicle routing problem (VRP), ignoring the transferable knowledge across different VRP variants. This paper proposes the cross-problem learning to assist heuristics training for different downstream VRP variants. Particularly, we modularize neural architectures for complex VRPs into 1) the backbone Transformer for tackling the travelling salesman problem (TSP), and 2) the additional lightweight modules for processing problem-specific features in complex VRPs. Accordingly, we propose to pre-train the backbone Transformer for TSP, and then apply it in the process of fine-tuning the Transformer models for each target VRP variant. On the one hand, we fully fine-tune the trained backbone Transformer and problem-specific modules simultaneously. On the other hand, we only fine-tune small adapter networks along with the modules, keeping the backbone Transformer still. Extensive experiments on typical VRPs substantiate that 1) the full fine-tuning achieves significantly better performance than the one trained from scratch, and 2) the adapter-based fine-tuning also delivers comparable performance while being notably parameter-efficient. Furthermore, we empirically demonstrate the favorable effect of our method in terms of cross-distribution application and versatility.

S-REINFORCE: A Neuro-Symbolic Policy Gradient Approach for Interpretable Reinforcement Learning

This paper presents a novel RL algorithm, S-REINFORCE, which is designed to generate interpretable policies for dynamic decision-making tasks. The proposed algorithm leverages two types of function approximators, namely Neural Network (NN) and Symbolic Regressor (SR), to produce numerical and symbolic policies, respectively. The NN component learns to generate a numerical probability distribution over the possible actions using a policy gradient, while the SR component captures the functional form that relates the associated states with the action probabilities. The SR-generated policy expressions are then utilized through importance sampling to improve the rewards received during the learning process. We have tested the proposed S-REINFORCE algorithm on various dynamic decision-making problems with low and high dimensional action spaces, and the results demonstrate its effectiveness and impact in achieving interpretable solutions. By leveraging the strengths of both NN and SR, S-REINFORCE produces policies that are not only well-performing but also easy to interpret, making it an ideal choice for real-world applications where transparency and causality are crucial.

Robust Representation Learning with Self-Distillation for Domain Generalization

Domain generalization is a challenging problem in machine learning, where the goal is to train a model that can generalize well to unseen target domains without prior knowledge of these domains. Despite the recent success of deep neural networks, there remains a lack of effective methods for domain generalization using vision transformers. In this paper, we propose a novel domain generalization technique called Robust Representation Learning with Self-Distillation (RRLD) that utilizes a combination of i) intermediate-block self-distillation and ii) augmentation-guided self-distillation to improve the generalization capabilities of transformer-based models on unseen domains. This approach enables the network to learn robust and general features that are invariant to different augmentations and domain shifts while effectively mitigating overfitting to source domains. To evaluate the effectiveness of our proposed method, we perform extensive experiments on PACS [1] and OfficeHome [2] benchmark datasets, as well as a real-world wafer semiconductor defect dataset [3]. Our results demonstrate that RRLD achieves robust and accurate generalization performance. We observe an improvement in the range of 0.3% to 2.3% over the state-of-the-art on the three datasets.

Latent Preserving Generative Adversarial Network for Imbalance classification

Many real-world classification problems have imbalanced frequency of class labels; a well-known issue known as the "class imbalance" problem. Classic classification algorithms tend to be biased towards the majority class, leaving the classifier vulnerable to misclassification of the minority class. While the literature is rich with methods to fix this problem, as the dimensionality of the problem increases, many of these methods do not scale-up and the cost of running them become prohibitive. In this paper, we present an end-to-end deep generative classifier. We propose a domain-constraint autoencoder to preserve the latent-space as prior for a generator, which is then used to play an adversarial game with two other deep networks, a discriminator and a classifier. Extensive experiments are carried out on three different multi-class imbalanced problems and a comparison with state-of-the-art methods. Experimental results confirmed the superiority of our method over popular algorithms in handling high-dimensional imbalanced classification problems. Our code is available on https://github.com/TanmDL/SLPPL-GAN.

Does Adversarial Oversampling Help us?

Traditional oversampling methods are generally employed to handle class imbalance in datasets. This oversampling approach is independent of the classifier; thus, it does not offer an end-to-end solution. To overcome this, we propose a three-player adversarial game-based end-to-end method, where a domain-constraints mixture of generators, a discriminator, and a multi-class classifier are used. Rather than adversarial minority oversampling, we propose an adversarial oversampling (AO) and a data-space oversampling (DO) approach. In AO, the generator updates by fooling both the classifier and discriminator, however, in DO, it updates by favoring the classifier and fooling the discriminator. While updating the classifier, it considers both the real and synthetically generated samples in AO. But, in DO, it favors the real samples and fools the subset class-specific generated samples. To mitigate the biases of a classifier towards the majority class, minority samples are over-sampled at a fractional rate. Such implementation is shown to provide more robust classification boundaries. The effectiveness of our proposed method has been validated with high-dimensional, highly imbalanced and large-scale multi-class tabular datasets. The results as measured by average class specific accuracy (ACSA) clearly indicate that the proposed method provides better classification accuracy (improvement in the range of 0.7% to 49.27%) as compared to the baseline classifier.

DO-GAN: A Double Oracle Framework for Generative Adversarial Networks

In this paper, we propose a new approach to train Generative Adversarial Networks (GANs) where we deploy a double-oracle framework using the generator and discriminator oracles. GAN is essentially a two-player zero-sum game between the generator and the discriminator. Training GANs is challenging as a pure Nash equilibrium may not exist and even finding the mixed Nash equilibrium is difficult as GANs have a large-scale strategy space. In DO-GAN, we extend the double oracle framework to GANs. We first generalize the players' strategies as the trained models of generator and discriminator from the best response oracles. We then compute the meta-strategies using a linear program. For scalability of the framework where multiple generators and discriminator best responses are stored in the memory, we propose two solutions: 1) pruning the weakly-dominated players' strategies to keep the oracles from becoming intractable; 2) applying continual learning to retain the previous knowledge of the networks. We apply our framework to established GAN architectures such as vanilla GAN, Deep Convolutional GAN, Spectral Normalization GAN and Stacked GAN. Finally, we conduct experiments on MNIST, CIFAR-10 and CelebA datasets and show that DO-GAN variants have significant improvements in both subjective qualitative evaluation and quantitative metrics, compared with their respective GAN architectures.

Inverse design of crystals using generalized invertible crystallographic representation

Deep learning has fostered many novel applications in materials informatics. However, the inverse design of inorganic crystals, $\textit{i.e.}$ generating new crystal structure with targeted properties, remains a grand challenge. An important ingredient for such generative models is an invertible representation that accesses the full periodic table. This is challenging due to limited data availability and the complexity of 3D periodic crystal structures. In this paper, we present a generalized invertible representation that encodes the crystallographic information into the descriptors in both real space and reciprocal space. Combining with a generative variational autoencoder (VAE), a wide range of crystallographic structures and chemistries with desired properties can be inverse-designed. We show that our VAE model predicts novel crystal structures that do not exist in the training and test database (Materials Project) with targeted formation energies and band gaps. We validate those predicted crystals by first-principles calculations. Finally, to design solids with practical applications, we address the sparse label problem by building a semi-supervised VAE and demonstrate its successful prediction of unique thermoelectric materials