In this paper, we propose Attention Based Decomposition Network (ABD-Net), for point cloud decomposition into basic geometric shapes namely, plane, sphere, cone and cylinder. We show improved performance of 3D object classification using attention features based on primitive shapes in point clouds. Point clouds, being the simple and compact representation of 3D objects have gained increasing popularity. They demand robust methods for feature extraction due to unorderness in point sets. In ABD-Net the proposed Local Proximity Encapsulator captures the local geometric variations along with spatial encoding around each point from the input point sets. The encapsulated local features are further passed to proposed Attention Feature Encoder to learn basic shapes in point cloud. Attention Feature Encoder models geometric relationship between the neighborhoods of all the points resulting in capturing global point cloud information. We demonstrate the results of our proposed ABD-Net on ANSI mechanical component and ModelNet40 datasets. We also demonstrate the effectiveness of ABD-Net over the acquired attention features by improving the performance of 3D object classification on ModelNet40 benchmark dataset and compare them with state-of-the-art techniques.

However, Our contributions in this work are: instead of computing the curvature matrix, we show that, under some regularity conditions, - We show that under some regularity conditions, a diagonal the Laplace approximation can be easily constructed Laplace approximation can be constructed without using the gradient second moment. This computing anything besides what is already being quantity is already estimated by many exponential computed by widely used optimizers; moving average variants of Adagrad such as Adam and RMSprop, but is traditionally discarded - We qualitatively compare this approximation with after training. We show that our method (L2M) methods such as deep ensembles (Lakshminarayanan does not require changes in models or optimization, et al., 2017), MC Dropout (Gal & Ghahramani, 2016), can be implemented in a few lines of code Hamiltonian Monte Carlo (HMC) (Cobb & Jalaian, to yield reasonable results, and it does not require 2020), among others; any extra computational steps besides what is already - We also show that our approximation is orthogonal being computed by optimizers, without introducing to methods such as ensembling (Lakshminarayanan any new hyperparameter. We hope our et al., 2017) and does not require changing training method can open new research directions on using procedures, estimating new quantities, or adding new quantities already computed by optimizers for hyperparameters.

For machine learning systems to be reliable, we must understand their performance in unseen, out-of-distribution environments. In this paper, we empirically show that out-of-distribution performance is strongly correlated with in-distribution performance for a wide range of models and distribution shifts. Specifically, we demonstrate strong correlations between in-distribution and out-of-distribution performance on variants of CIFAR-10 & ImageNet, a synthetic pose estimation task derived from YCB objects, satellite imagery classification in FMoW-WILDS, and wildlife classification in iWildCam-WILDS. The strong correlations hold across model architectures, hyperparameters, training set size, and training duration, and are more precise than what is expected from existing domain adaptation theory. To complete the picture, we also investigate cases where the correlation is weaker, for instance some synthetic distribution shifts from CIFAR-10-C and the tissue classification dataset Camelyon17-WILDS. Finally, we provide a candidate theory based on a Gaussian data model that shows how changes in the data covariance arising from distribution shift can affect the observed correlations.

We show that many machine-learning algorithms are specific instances of a single algorithm called the Bayesian learning rule. The rule, derived from Bayesian principles, yields a wide-range of algorithms from fields such as optimization, deep learning, and graphical models. This includes classical algorithms such as ridge regression, Newton's method, and Kalman filter, as well as modern deep-learning algorithms such as stochastic-gradient descent, RMSprop, and Dropout. The key idea in deriving such algorithms is to approximate the posterior using candidate distributions estimated by using natural gradients. Different candidate distributions result in different algorithms and further approximations to natural gradients give rise to variants of those algorithms. Our work not only unifies, generalizes, and improves existing algorithms, but also helps us design new ones.

Ensembles of CNN models trained with different seeds (also known as Deep Ensembles) are known to achieve superior performance over a single copy of the CNN. Neural Ensemble Search (NES) can further boost performance by adding architectural diversity. However, the scope of NES remains prohibitive under limited computational resources. In this work, we extend NES to multi-headed ensembles, which consist of a shared backbone attached to multiple prediction heads. Unlike Deep Ensembles, these multi-headed ensembles can be trained end to end, which enables us to leverage one-shot NAS methods to optimize an ensemble objective. With extensive empirical evaluations, we demonstrate that multi-headed ensemble search finds robust ensembles 3 times faster, while having comparable performance to other ensemble search methods, in both predictive performance and uncertainty calibration.

The Fisher information is one of the most fundamental concepts in statistical machine learning. Intuitively, it measures the amount of information carried by a single random observation when the underlying model varies along certain directions in the parameter space: if such a variation does not change the underlying model, then a corresponding observation contains zero (Fisher) information and is non-informative regarding the varied parameter. Parameter estimation is impossible in this case. Otherwise, if the variation significantly changes the model and has large information, then an observation is informative and the parameter estimation can be more efficient as compared to parameters with small Fisher information. In machine learning, this basic concept is useful for defining intrinsic structures of the parameter space, measuring model complexity, and performing gradient-based optimization.

Tensor Networks, a numerical tool originally designed for simulating quantum many-body systems, have recently been applied to solve Machine Learning problems. Exploiting a tree tensor network, we apply a quantum-inspired machine learning technique to a very important and challenging big data problem in high energy physics: the analysis and classification of data produced by the Large Hadron Collider at CERN. In particular, we present how to effectively classify so-called b-jets, jets originating from b-quarks from proton-proton collisions in the LHCb experiment, and how to interpret the classification results. We exploit the Tensor Network approach to select important features and adapt the network geometry based on information acquired in the learning process. Finally, we show how to adapt the tree tensor network to achieve optimal precision or fast response in time without the need of repeating the learning process. These results pave the way to the implementation of high-frequency real-time applications, a key ingredient needed among others for current and future LHCb event classification able to trigger events at the tens of MHz scale.

Despite the groundbreaking successes of neural networks, contemporary models require extensive training with massive datasets and exhibit poor out-of-sample generalization. One proposed solution is to build systematicity and domain-specific constraints into the model, echoing the tenets of classical, symbolic cognitive architectures. In this paper, we consider the limitations of this approach by examining human adults' ability to learn an abstract reasoning task from a brief instructional tutorial and explanatory feedback for incorrect responses, demonstrating that human learning dynamics and ability to generalize outside the range of the training examples differ drastically from those of a representative neural network model, and that the model is brittle to changes in features not anticipated by its authors. We present further evidence from human data that the ability to consistently solve the puzzles was associated with education, particularly basic mathematics education, and with the ability to provide a reliably identifiable, valid description of the strategy used. We propose that rapid learning and systematic generalization in humans may depend on a gradual, experience-dependent process of learning-to-learn using instructions and explanations to guide the construction of explicit abstract rules that support generalizable inferences.

In this paper, we propose an end-to-end lifelong learning mixture of experts. Each expert is implemented by a Variational Autoencoder (VAE). The experts in the mixture system are jointly trained by maximizing a mixture of individual component evidence lower bounds (MELBO) on the log-likelihood of the given training samples. The mixing coefficients in the mixture, control the contributions of each expert in the goal representation. These are sampled from a Dirichlet distribution whose parameters are determined through non-parametric estimation during lifelong learning. The model can learn new tasks fast when these are similar to those previously learnt. The proposed Lifelong mixture of VAE (L-MVAE) expands its architecture with new components when learning a completely new task. After the training, our model can automatically determine the relevant expert to be used when fed with new data samples. This mechanism benefits both the memory efficiency and the required computational cost as only one expert is used during the inference. The L-MVAE inference model is able to perform interpolation in the joint latent space across the data domains associated with different tasks and is shown to be efficient for disentangled learning representation.

A unique cognitive capability of humans consists in their ability to acquire new knowledge and skills from a sequence of experiences. Meanwhile, artificial intelligence systems are good at learning only the last given task without being able to remember the databases learnt in the past. We propose a novel lifelong learning methodology by employing a Teacher-Student network framework. While the Student module is trained with a new given database, the Teacher module would remind the Student about the information learnt in the past. The Teacher, implemented by a Generative Adversarial Network (GAN), is trained to preserve and replay past knowledge corresponding to the probabilistic representations of previously learn databases. Meanwhile, the Student module is implemented by a Variational Autoencoder (VAE) which infers its latent variable representation from both the output of the Teacher module as well as from the newly available database. Moreover, the Student module is trained to capture both continuous and discrete underlying data representations across different domains. The proposed lifelong learning framework is applied in supervised, semi-supervised and unsupervised training. The code is available~: \url{https://github.com/dtuzi123/Lifelong-Teacher-Student-Network-Learning}