arXiv subjects

Xiuwen Liu

Publications and source records attributed to Xiuwen Liu.

34 records · Page 2Linked to original sources

arXiv2024

Learning Regularities from Data using Spiking Functions: A Theory

Deep neural networks trained in an end-to-end manner are proven to be efficient in a wide range of machine learning tasks. However, there is one drawback of end-to-end learning: The learned features and information are implicitly represented in neural network parameters, which cannot be used as regularities, concepts or knowledge to explicitly represent the data probability distribution. To resolve this issue, we propose in this paper a new machine learning theory, which defines in mathematics what are regularities. Briefly, regularities are concise representations of the non-random features, or 'non-randomness' in the data probability distribution. Combining this with information theory, we claim that regularities can also be regarded as a small amount of information encoding a large amount of information. Our theory is based on spiking functions. That is, if a function can react to, or spike on specific data samples more frequently than random noise inputs, we say that such a function discovers non-randomness from the data distribution. Also, we say that the discovered non-randomness is encoded into regularities if the function is simple enough. Our theory also discusses applying multiple spiking functions to the same data distribution. In this process, we claim that the 'best' regularities, or the optimal spiking functions, are those who can capture the largest amount of information from the data distribution, and then encode the captured information in the most concise way. Theorems and hypotheses are provided to describe in mathematics what are 'best' regularities and optimal spiking functions. Finally, we propose a machine learning approach, which can potentially obtain the optimal spiking functions regarding the given dataset in practice.

cs.LG↗

arXiv2024

Intriguing Differences Between Zero-Shot and Systematic Evaluations of Vision-Language Transformer Models

Transformer-based models have dominated natural language processing and other areas in the last few years due to their superior (zero-shot) performance on benchmark datasets. However, these models are poorly understood due to their complexity and size. While probing-based methods are widely used to understand specific properties, the structures of the representation space are not systematically characterized; consequently, it is unclear how such models generalize and overgeneralize to new inputs beyond datasets. In this paper, based on a new gradient descent optimization method, we are able to explore the embedding space of a commonly used vision-language model. Using the Imagenette dataset, we show that while the model achieves over 99\% zero-shot classification performance, it fails systematic evaluations completely. Using a linear approximation, we provide a framework to explain the striking differences. We have also obtained similar results using a different model to support that our results are applicable to other transformer models with continuous inputs. We also propose a robust way to detect the modified images.

cs.CV↗

arXiv2024

Intriguing Equivalence Structures of the Embedding Space of Vision Transformers

Pre-trained large foundation models play a central role in the recent surge of artificial intelligence, resulting in fine-tuned models with remarkable abilities when measured on benchmark datasets, standard exams, and applications. Due to their inherent complexity, these models are not well understood. While small adversarial inputs to such models are well known, the structures of the representation space are not well characterized despite their fundamental importance. In this paper, using the vision transformers as an example due to the continuous nature of their input space, we show via analyses and systematic experiments that the representation space consists of large piecewise linear subspaces where there exist very different inputs sharing the same representations, and at the same time, local normal spaces where there are visually indistinguishable inputs having very different representations. The empirical results are further verified using the local directional estimations of the Lipschitz constants of the underlying models. Consequently, the resulting representations change the results of downstream models, and such models are subject to overgeneralization and with limited semantically meaningful generalization capability.

cs.CV↗

arXiv2023

Nonlinear Correct and Smooth for Semi-Supervised Learning

Graph-based semi-supervised learning (GSSL) has been used successfully in various applications. Existing methods leverage the graph structure and labeled samples for classification. Label Propagation (LP) and Graph Neural Networks (GNNs) both iteratively pass messages on graphs, where LP propagates node labels through edges and GNN aggregates node features from the neighborhood. Recently, combining LP and GNN has led to improved performance. However, utilizing labels and features jointly in higher-order graphs has not been explored. Therefore, we propose Nonlinear Correct and Smooth (NLCS), which improves the existing post-processing approach by incorporating non-linearity and higher-order representation into the residual propagation to handle intricate node relationships effectively. Systematic evaluations show that our method achieves remarkable average improvements of 13.71% over base prediction and 2.16% over the state-of-the-art post-processing method on six commonly used datasets. Comparisons and analyses show our method effectively utilizes labels and features jointly in higher-order graphs to resolve challenging graph relationships.

cs.LG↗

arXiv2023

Automatic Historical Stock Price Dataset Generation Using Python

With the dynamic political and economic environments, the ever-changing stock markets generate large amounts of data daily. Acquiring up-to-date data is crucial to enhancing predictive precision in stock price behavior studies. However, preparing the dataset manually can be challenging and time-demanding. The stock market analysis usually revolves around specific indices such as S&P500, Nasdaq, Dow Jones, the New York Stock Exchange (NYSE), etc. It is necessary to analyze all the companies of any particular index. While raw data are accessible from diverse financial websites, these resources are tailored for individual company data retrieval and there is a big gap between what is available and what is needed to generate large datasets. Python emerges as a valuable tool for comprehensively collecting all constituent stocks within a given index. While certain online sources offer code snippets for limited dataset generation, a comprehensive and unified script is yet to be developed and publicly available. Therefore, we present a comprehensive and consolidated code resource that facilitates the extraction of updated datasets for any particular time period and for any specific stock market index and closes the gap. The code is available at https://github.com/amp1590/automatic_stock_data_collection.

cs.CE↗

arXiv2023

Understanding the Spectral Bias of Coordinate Based MLPs Via Training Dynamics

Spectral bias is an important observation of neural network training, stating that the network will learn a low frequency representation of the target function before converging to higher frequency components. This property is interesting due to its link to good generalization in over-parameterized networks. However, in low dimensional settings, a severe spectral bias occurs that obstructs convergence to high frequency components entirely. In order to overcome this limitation, one can encode the inputs using a high frequency sinusoidal encoding. Previous works attempted to explain this phenomenon using Neural Tangent Kernel (NTK) and Fourier analysis. However, NTK does not capture real network dynamics, and Fourier analysis only offers a global perspective on the network properties that induce this bias. In this paper, we provide a novel approach towards understanding spectral bias by directly studying ReLU MLP training dynamics. Specifically, we focus on the connection between the computations of ReLU networks (activation regions), and the speed of gradient descent convergence. We study these dynamics in relation to the spatial information of the signal to understand how they influence spectral bias. We then use this formulation to study the severity of spectral bias in low dimensional settings, and how positional encoding overcomes this.

cs.LG↗

arXiv2023

Improved Stock Price Movement Classification Using News Articles Based on Embeddings and Label Smoothing

Stock price movement prediction is a challenging and essential problem in finance. While it is well established in modern behavioral finance that the share prices of related stocks often move after the release of news via reactions and overreactions of investors, how to capture the relationships between price movements and news articles via quantitative models is an active area research; existing models have achieved success with variable degrees. In this paper, we propose to improve stock price movement classification using news articles by incorporating regularization and optimization techniques from deep learning. More specifically, we capture the dependencies between news articles and stocks through embeddings and bidirectional recurrent neural networks as in recent models. We further incorporate weight decay, batch normalization, dropout, and label smoothing to improve the generalization of the trained models. To handle high fluctuations of validation accuracy of batch normalization, we propose dual-phase training to realize the improvements reliably. Our experimental results on a commonly used dataset show significant improvements, achieving average accuracy of 80.7% on the test set, which is more than 10.0% absolute improvement over existing models. Our ablation studies show batch normalization and label smoothing are most effective, leading to 6.0% and 3.4% absolute improvement, respectively on average.

cs.LG↗

arXiv2022

Dense Embeddings Preserving the Semantic Relationships in WordNet

In this paper, we provide a novel way to generate low dimensional vector embeddings for the noun and verb synsets in WordNet, where the hypernym-hyponym relationship is preserved in the embeddings. We call this embedding the Sense Spectrum (and Sense Spectra for embeddings). In order to create suitable labels for the training of sense spectra, we designed a new similarity measurement for noun and verb synsets in WordNet. We call this similarity measurement the Hypernym Intersection Similarity (HIS), since it compares the common and unique hypernyms between two synsets. Our experiments show that on the noun and verb pairs of the SimLex-999 dataset, HIS outperforms the three similarity measurements in WordNet. Moreover, to the best of our knowledge, the sense spectra provide the first dense synset embeddings that preserve the semantic relationships in WordNet.

cs.CL↗

arXiv2020

An Analysis on the Learning Rules of the Skip-Gram Model

To improve the generalization of the representations for natural language processing tasks, words are commonly represented using vectors, where distances among the vectors are related to the similarity of the words. While word2vec, the state-of-the-art implementation of the skip-gram model, is widely used and improves the performance of many natural language processing tasks, its mechanism is not yet well understood. In this work, we derive the learning rules for the skip-gram model and establish their close relationship to competitive learning. In addition, we provide the global optimal solution constraints for the skip-gram model and validate them by experimental results.

cs.CL↗

arXiv2019

Towards Quantifying Intrinsic Generalization of Deep ReLU Networks

Understanding the underlying mechanisms that enable the empirical successes of deep neural networks is essential for further improving their performance and explaining such networks. Towards this goal, a specific question is how to explain the "surprising" behavior of the same over-parametrized deep neural networks that can generalize well on real datasets and at the same time "memorize" training samples when the labels are randomized. In this paper, we demonstrate that deep ReLU networks generalize from training samples to new points via piece-wise linear interpolation. We provide a quantified analysis on the generalization ability of a deep ReLU network: Given a fixed point $\mathbf{x}$ and a fixed direction in the input space $\mathcal{S}$, there is always a segment such that any point on the segment will be classified the same as the fixed point $\mathbf{x}$. We call this segment the $generalization \ interval$. We show that the generalization intervals of a ReLU network behave similarly along pairwise directions between samples of the same label in both real and random cases on the MNIST and CIFAR-10 datasets. This result suggests that the same interpolation mechanism is used in both cases. Additionally, for datasets using real labels, such networks provide a good approximation of the underlying manifold in the data, where the changes are much smaller along tangent directions than along normal directions. On the other hand, however, for datasets with random labels, generalization intervals along mid-lines of triangles with the same label are much smaller than those on the datasets with real labels, suggesting different behaviors along other directions. Our systematic experiments demonstrate for the first time that such deep neural networks generalize through the same interpolation and explain the differences between their performance on datasets with real and random labels.

cs.LG↗

arXiv2019

Consensus-based Interpretable Deep Neural Networks with Application to Mortality Prediction

Deep neural networks have achieved remarkable success in various challenging tasks. However, the black-box nature of such networks is not acceptable to critical applications, such as healthcare. In particular, the existence of adversarial examples and their overgeneralization to irrelevant, out-of-distribution inputs with high confidence makes it difficult, if not impossible, to explain decisions by such networks. In this paper, we analyze the underlying mechanism of generalization of deep neural networks and propose an ($n$, $k$) consensus algorithm which is insensitive to adversarial examples and can reliably reject out-of-distribution samples. Furthermore, the consensus algorithm is able to improve classification accuracy by using multiple trained deep neural networks. To handle the complexity of deep neural networks, we cluster linear approximations of individual models and identify highly correlated clusters among different models to capture feature importance robustly, resulting in improved interpretability. Motivated by the importance of building accurate and interpretable prediction models for healthcare, our experimental results on an ICU dataset show the effectiveness of our algorithm in enhancing both the prediction accuracy and the interpretability of deep neural network models on one-year patient mortality prediction. In particular, while the proposed method maintains similar interpretability as conventional shallow models such as logistic regression, it improves the prediction accuracy significantly.

cs.LG↗

arXiv2019

Next-Generation High-Resolution Vector-Borne Disease Risk Assessment

Vector-borne diseases cause more than 1 million deaths annually. Estimates of epidemic risk at high spatial resolutions can enable effective public health interventions. Our goal is to identify the risk of importation of such diseases into vulnerable cities at the granularity of neighborhoods. Conventional models cannot achieve such spatial resolution, especially in real-time. Besides, they lack real-time data on demographic heterogeneity, which is vital for accurate risk estimation. Social media, such as Twitter, promise data from which demographic and spatial information could be inferred in real-time. On the other hand, such data can be noisy and inaccurate. Our novel approach leverages Twitter data, using machine learning techniques at multiple spatial scales to overcome its limitations, to deliver results at the desired resolution. We validate our method against the Zika outbreak in Florida in 2016. Our main contribution lies in proposing a novel approach that uses machine learning on social media data to identify the risk of vector-borne disease importation at a sufficiently fine spatial resolution to permit effective intervention. This will lead to a new generation of epidemic risk assessment models, promising to transform public health by identifying specific locations for targeted intervention.

cs.SI↗

arXiv2019

High-resolution home location prediction from tweets using deep learning with dynamic structure

Timely and high-resolution estimates of the home locations of a sufficiently large subset of the population are critical for effective disaster response and public health intervention, but this is still an open problem. Conventional data sources, such as census and surveys, have a substantial time lag and cannot capture seasonal trends. Recently, social media data has been exploited to address this problem by leveraging its large user-base and real-time nature. However, inherent sparsity and noise, along with large estimation uncertainty in home locations, have limited their effectiveness. Consequently, much of previous research has aimed only at a coarse spatial resolution, with accuracy being limited for high-resolution methods. In this paper, we develop a deep-learning solution that uses a two-phase dynamic structure to deal with sparse and noisy social media data. In the first phase, high recall is achieved using a random forest, producing more balanced home location candidates. Then two deep neural networks are used to detect home locations with high accuracy. We obtained over 90% accuracy for large subsets on a commonly used dataset. Compared to other high-resolution methods, our approach yields up to 60% error reduction by reducing high-resolution home prediction error from over 21% to less than 8%. Systematic comparisons show that our method gives the highest accuracy both for the entire sample and for subsets. Evaluation on a real-world public health problem further validates the effectiveness of our approach.

cs.SI↗

arXiv2019

Overfitting Mechanism and Avoidance in Deep Neural Networks

Assisted by the availability of data and high performance computing, deep learning techniques have achieved breakthroughs and surpassed human performance empirically in difficult tasks, including object recognition, speech recognition, and natural language processing. As they are being used in critical applications, understanding underlying mechanisms for their successes and limitations is imperative. In this paper, we show that overfitting, one of the fundamental issues in deep neural networks, is due to continuous gradient updating and scale sensitiveness of cross entropy loss. By separating samples into correctly and incorrectly classified ones, we show that they behave very differently, where the loss decreases in the correct ones and increases in the incorrect ones. Furthermore, by analyzing dynamics during training, we propose a consensus-based classification algorithm that enables us to avoid overfitting and significantly improve the classification accuracy especially when the number of training samples is limited. As each trained neural network depends on extrinsic factors such as initial values as well as training data, requiring consensus among multiple models reduces extrinsic factors substantially; for statistically independent models, the reduction is exponential. Compared to ensemble algorithms, the proposed algorithm avoids overgeneralization by not classifying ambiguous inputs. Systematic experimental results demonstrate the effectiveness of the proposed algorithm. For example, using only 1000 training samples from MNIST dataset, the proposed algorithm achieves 95% accuracy, significantly higher than any of the individual models, with 90% of the test samples classified.

cs.LG↗

arXiv2017

Land Cover Classification from Multi-temporal, Multi-spectral Remotely Sensed Imagery using Patch-Based Recurrent Neural Networks

Sustainability of the global environment is dependent on the accurate land cover information over large areas. Even with the increased number of satellite systems and sensors acquiring data with improved spectral, spatial, radiometric and temporal characteristics and the new data distribution policy, most existing land cover datasets were derived from a pixel-based single-date multi-spectral remotely sensed image with low accuracy. To improve the accuracy, the bottleneck is how to develop an accurate and effective image classification technique. By incorporating and utilizing the complete multi-spectral, multi-temporal and spatial information in remote sensing images and considering their inherit spatial and sequential interdependence, we propose a new patch-based RNN (PB-RNN) system tailored for multi-temporal remote sensing data. The system is designed by incorporating distinctive characteristics in multi-temporal remote sensing data. In particular, it uses multi-temporal-spectral-spatial samples and deals with pixels contaminated by clouds/shadow present in the multi-temporal data series. Using a Florida Everglades ecosystem study site covering an area of 771 square kilo-meters, the proposed PB-RNN system has achieved a significant improvement in the classification accuracy over pixel-based RNN system, pixel-based single-imagery NN system, pixel-based multi-images NN system, patch-based single-imagery NN system and patch-based multi-images NN system. For example, the proposed system achieves 97.21% classification accuracy while a pixel-based single-imagery NN system achieves 64.74%. By utilizing methods like the proposed PB-RNN one, we believe that much more accurate land cover datasets can be produced over large areas efficiently.

cs.CV↗

arXiv2015

Transformed Residual Quantization for Approximate Nearest Neighbor Search

The success of product quantization (PQ) for fast nearest neighbor search depends on the exponentially reduced complexities of both storage and computation with respect to the codebook size. Recent efforts have been focused on employing sophisticated optimization strategies, or seeking more effective models. Residual quantization (RQ) is such an alternative that holds the same property as PQ in terms of the aforementioned complexities. In addition to being a direct replacement of PQ, hybrids of PQ and RQ can yield more gains for approximate nearest neighbor search. This motivated us to propose a novel approach to optimizing RQ and the related hybrid models. With an observation of the general randomness increase in a residual space, we propose a new strategy that jointly learns a local transformation per residual cluster with an ultimate goal to reduce overall quantization errors. We have shown that our approach can achieve significantly better accuracy on nearest neighbor search than both the original and the optimized PQ on several very large scale benchmarks.

cs.CV↗