In extreme classification settings, embedding-based neural network models are currently not competitive with sparse linear and tree-based methods in terms of accuracy. Most prior works attribute this poor performance to the low-dimensional bottleneck in embedding-based methods. In this paper, we demonstrate that theoretically there is no limitation to using low-dimensional embedding-based methods, and provide experimental evidence that overfitting is the root cause of the poor performance of embedding-based methods. These findings motivate us to investigate novel data augmentation and regularization techniques to mitigate overfitting. To this end, we propose GLaS, a new regularizer for embedding-based neural network approaches. It is a natural generalization from the graph Laplacian and spread-out regularizers, and empirically it addresses the drawback of each regularizer alone when applied to the extreme classification setup. With the proposed techniques, we attain or improve upon the state-of-the-art on most widely tested public extreme classification datasets with hundreds of thousands of labels.

Learning of low dimensional structure in multidimensional data is a canonical problem in machine learning. One common approach is to suppose that the observed data are close to a lower-dimensional smooth manifold. There are a rich variety of manifold learning methods available, which allow mapping of data points to the manifold. However, there is a clear lack of probabilistic methods that allow learning of the manifold along with the generative distribution of the observed data. The best attempt is the Gaussian process latent variable model (GP-L VM), but identifiability issues lead to poor performance. We solve these issues by proposing a novel Coulomb repulsive process (Corp) for locations of points on the manifold, inspired by physical models of electrostatic interactions among particles. Combining this process with a GP prior for the mapping function yields a novel electrostatic GP (electroGP) process. Focusing on the simple case of a one-dimensional manifold, we develop efficient inference algorithms, and illustrate substantially improved performance in a variety of experiments including filling in missing frames in video.

In extreme classification settings, embedding-based neural network models are currently not competitive with sparse linear and tree-based methods in terms of accuracy. Most prior works attribute this poor performance to the low-dimensional bottleneck in embedding-based methods. In this paper, we demonstrate that theoretically there is no limitation to using low-dimensional embedding-based methods, and provide experimental evidence that overfitting is the root cause of the poor performance of embedding-based methods. These findings motivate us to investigate novel data augmentation and regularization techniques to mitigate overfitting. To this end, we propose GLaS, a new regularizer for embedding-based neural network approaches.

In the prior literature, they cited the low dimensional embedding methods is the reason of the poor performance of the embedding based methods. In this paper, the author proposed that the final score vector for the labels actually generated by highly non-linear transformation such as thresholding the scores. Thus it is not clear if the low-rank structure of the score vectors directly cause the low-rank on the label vectors. Furthermore, the author uses a simple neural network to mimic the low-dimensional embedding can attain near-perfect training accuracy but generalize poorly and suggesting that overfitting is the root cause of the poor performance of the embedding based methods. This is the first contribution of the paper which breaks the glass ceiling of embedding based methods.

In extreme classification settings, embedding-based neural network models are currently not competitive with sparse linear and tree-based methods in terms of accuracy. Most prior works attribute this poor performance to the low-dimensional bottleneck in embedding-based methods. In this paper, we demonstrate that theoretically there is no limitation to using low-dimensional embedding-based methods, and provide experimental evidence that overfitting is the root cause of the poor performance of embedding-based methods. These findings motivate us to investigate novel data augmentation and regularization techniques to mitigate overfitting. To this end, we propose GLaS, a new regularizer for embedding-based neural network approaches.

Microsoft outperformed analyst predictions in its latest quarterly earnings report, revealing on Tuesday that its revenue was up 15% year-over-year. But growth of the company's closely watched Azure cloud computing services failed to meet expectations and shares in Microsoft fell as much as 7% in after-hours trading. The company was expected to report steady growth in its fourth quarter earnings report, mostly on the back of its cloud services. Revenue from those services grew 29%, lower than the 30% to 31% that analysts predicted, resulting in a sell-off that exacerbates big tech's recent market woes. In Microsoft's earnings report, Satya Nadella, the CEO, sought to bolster confidence in the company's performance. "Our strong performance this fiscal year speaks both to our innovation and to the trust customers continue to place in Microsoft," said Nadella in the earnings statement.

Learning of low dimensional structure in multidimensional data is a canonical problem in machine learning. One common approach is to suppose that the observed data are close to a lower-dimensional smooth manifold. There are a rich variety of manifold learning methods available, which allow mapping of data points to the manifold. However, there is a clear lack of probabilistic methods that allow learning of the manifold along with the generative distribution of the observed data. The best attempt is the Gaussian process latent variable model (GP-LVM), but identifiability issues lead to poor performance. We solve these issues by proposing a novel Coulomb repulsive process (Corp) for locations of points on the manifold, inspired by physical models of electrostatic interactions among particles. Combining this process with a GP prior for the mapping function yields a novel electrostatic GP (electroGP) process. Focusing on the simple case of a one-dimensional manifold, we develop efficient inference algorithms, and illustrate substantially improved performance in a variety of experiments including filling in missing frames in video.

Small obstacles on the ground often lead to a fall when caught with commercial prosthetic feet. Despite some recently developed feet can actively control the ankle angle, for instance over slopes, their flat and rigid sole remains a cause of instability on uneven grounds. Soft robotic feet were recently proposed to tackle that issue; however, they lack consistent experimental validation. Therefore, this paper describes the experimental setup realized to test soft and rigid prosthetic feet with lower-limb prosthetic users. It includes a wooden walkway and differently shaped obstacles. It was preliminary validated with an able-bodied subject, the same subject walking on commercial prostheses through modified walking boots, and with a prosthetic user. They performed walking firstly on even ground, and secondly on even ground stepping on one of the obstacles. Results in terms of vertical ground reaction force and knee moments in both the sagittal and frontal planes show how the poor performance of commonly used prostheses is exacerbated in case of obstacles. The prosthetic user, indeed, noticeably relies on the sound leg to compensate for the stiff and unstable interaction of the prosthetic limb with the obstacle. Therefore, since the limitations of non-adaptive prosthetic feet in obstacle-dealing emerge from the experiments, as expected, this study justifies the use of the setup for investigating the performance of soft feet on uneven grounds and obstacle negotiation.

Interpreting machine learning models remains a challenge, hindering their adoption in clinical settings. This paper proposes leveraging Local Interpretable Model-Agnostic Explanations (LIME) to provide interpretable descriptions of black box classification models in high-stakes sepsis detection. By analyzing misclassified instances, significant features contributing to suboptimal performance are identified. The analysis reveals regions where the classifier performs poorly, allowing the calculation of error rates within these regions. This knowledge is crucial for cautious decision-making in sepsis detection and other critical applications. The proposed approach is demonstrated using the eICU dataset, effectively identifying and visualizing regions where the classifier underperforms. By enhancing interpretability, our method promotes the adoption of machine learning models in clinical practice, empowering informed decision-making and mitigating risks in critical scenarios.

Originally published on Towards AI the World's Leading AI and Technology News and Media Company. If you are building an AI-related product or service, we invite you to consider becoming an AI sponsor. At Towards AI, we help scale AI and technology startups. Let us help you unleash your technology to the masses. Are you tired of your machine learning models performing poorly on new data? Are you sick of seeing your model's validation accuracy skyrocket only to crash and burn on the test set?