arXiv subjects

Beidi Chen

Publications and source records attributed to Beidi Chen.

83 records · Page 5Linked to original sources

arXiv2021

Satellite Images and Deep Learning to Identify Discrepancy in Mailing Addresses with Applications to Census 2020 in Houston

The accuracy and completeness of population estimation would significantly impact the allocation of public resources. However, the current census paradigm experiences a non-negligible level of under-counting. Existing solutions to this problem by the Census Bureau is to increase canvassing efforts, which leads to expensive and inefficient usage of human resources. In this work, we argue that the existence of hidden multi-family households is a significant cause of under-counting. Accordingly, we introduce a low-cost but high-accuracy method that combines satellite imagery and deep learning technologies to identify hidden multi-family (HMF) households. With comprehensive knowledge of the HMF households, the efficiency and effectiveness of the decennial census could be vastly improved. An extensive experiment demonstrates that our approach can discover over 1800 undetected HMF in a single zipcode of the Houston area.

cs.CY↗

arXiv2021

Scatterbrain: Unifying Sparse and Low-rank Attention Approximation

Recent advances in efficient Transformers have exploited either the sparsity or low-rank properties of attention matrices to reduce the computational and memory bottlenecks of modeling long sequences. However, it is still challenging to balance the trade-off between model quality and efficiency to perform a one-size-fits-all approximation for different tasks. To better understand this trade-off, we observe that sparse and low-rank approximations excel in different regimes, determined by the softmax temperature in attention, and sparse + low-rank can outperform each individually. Inspired by the classical robust-PCA algorithm for sparse and low-rank decomposition, we propose Scatterbrain, a novel way to unify sparse (via locality sensitive hashing) and low-rank (via kernel feature map) attention for accurate and efficient approximation. The estimation is unbiased with provably low error. We empirically show that Scatterbrain can achieve 2.1x lower error than baselines when serving as a drop-in replacement in BigGAN image generation and pre-trained T2T-ViT. On a pre-trained T2T Vision transformer, even without fine-tuning, Scatterbrain can reduce 98% of attention memory at the cost of only 1% drop in accuracy. We demonstrate Scatterbrain for end-to-end training with up to 4 points better perplexity and 5 points better average accuracy than sparse or low-rank efficient transformers on language modeling and long-range-arena tasks.

cs.LG↗

arXiv2021

A Tale of Two Efficient and Informative Negative Sampling Distributions

Softmax classifiers with a very large number of classes naturally occur in many applications such as natural language processing and information retrieval. The calculation of full softmax is costly from the computational and energy perspective. There have been various sampling approaches to overcome this challenge, popularly known as negative sampling (NS). Ideally, NS should sample negative classes from a distribution that is dependent on the input data, the current parameters, and the correct positive class. Unfortunately, due to the dynamically updated parameters and data samples, there is no sampling scheme that is provably adaptive and samples the negative classes efficiently. Therefore, alternative heuristics like random sampling, static frequency-based sampling, or learning-based biased sampling, which primarily trade either the sampling cost or the adaptivity of samples per iteration are adopted. In this paper, we show two classes of distributions where the sampling scheme is truly adaptive and provably generates negative samples in near-constant time. Our implementation in C++ on CPU is significantly superior, both in terms of wall-clock time and accuracy, compared to the most optimized TensorFlow implementations of other popular negative sampling approaches on powerful NVIDIA V100 GPU.

cs.LG↗

arXiv2020

SOLAR: Sparse Orthogonal Learned and Random Embeddings

Dense embedding models are commonly deployed in commercial search engines, wherein all the document vectors are pre-computed, and near-neighbor search (NNS) is performed with the query vector to find relevant documents. However, the bottleneck of indexing a large number of dense vectors and performing an NNS hurts the query time and accuracy of these models. In this paper, we argue that high-dimensional and ultra-sparse embedding is a significantly superior alternative to dense low-dimensional embedding for both query efficiency and accuracy. Extreme sparsity eliminates the need for NNS by replacing them with simple lookups, while its high dimensionality ensures that the embeddings are informative even when sparse. However, learning extremely high dimensional embeddings leads to blow up in the model size. To make the training feasible, we propose a partitioning algorithm that learns such high dimensional embeddings across multiple GPUs without any communication. This is facilitated by our novel asymmetric mixture of Sparse, Orthogonal, Learned and Random (SOLAR) Embeddings. The label vectors are random, sparse, and near-orthogonal by design, while the query vectors are learned and sparse. We theoretically prove that our way of one-sided learning is equivalent to learning both query and label embeddings. With these unique properties, we can successfully train 500K dimensional SOLAR embeddings for the tasks of searching through 1.6M books and multi-label classification on the three largest public datasets. We achieve superior precision and recall compared to the respective state-of-the-art baselines for each of the tasks with up to 10 times faster speed.

cs.LG↗

arXiv2020

Angular Visual Hardness

Recent convolutional neural networks (CNNs) have led to impressive performance but often suffer from poor calibration. They tend to be overconfident, with the model confidence not always reflecting the underlying true ambiguity and hardness. In this paper, we propose angular visual hardness (AVH), a score given by the normalized angular distance between the sample feature embedding and the target classifier to measure sample hardness. We validate this score with an in-depth and extensive scientific study, and observe that CNN models with the highest accuracy also have the best AVH scores. This agrees with an earlier finding that state-of-art models improve on the classification of harder examples. We observe that the training dynamics of AVH is vastly different compared to the training loss. Specifically, AVH quickly reaches a plateau for all samples even though the training loss keeps improving. This suggests the need for designing better loss functions that can target harder examples more effectively. We also find that AVH has a statistically significant correlation with human visual hardness. Finally, we demonstrate the benefit of AVH to a variety of applications such as self-training for domain adaptation and domain generalization.

cs.LG↗

arXiv2020

Climbing the WOL: Training for Cheaper Inference

Efficient inference for wide output layers (WOLs) is an essential yet challenging task in large scale machine learning. Most approaches reduce this problem to approximate maximum inner product search (MIPS), which relies heavily on the observation that for a given model, ground truth labels correspond to logits of highest value during full model inference. However, such an assumption is restrictive in practice. In this paper, we argue that approximate MIPS subroutines, despite having sub-linear computation time, are sub-optimal because they are tailored for retrieving large inner products with high recall instead of retrieving the correct labels. With WOL, the labels often have moderate inner products, which makes approximate MIPS more challenging. We propose an alternative problem formulation, called Label Superior Sampling (LSS), where the objective is to tailor the system to ensure retrieval of the correct label. Accordingly, we propose a novel learned hash approach, which is significantly more efficient and sufficient for high inference accuracy than MIPS baselines. Our extensive evaluation indicates that LSS can match or even outperform full inference accuracy with around 5x speed up and 87% energy reduction.

cs.IR↗

arXiv2020

SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems

Deep Learning (DL) algorithms are the central focus of modern machine learning systems. As data volumes keep growing, it has become customary to train large neural networks with hundreds of millions of parameters to maintain enough capacity to memorize these volumes and obtain state-of-the-art accuracy. To get around the costly computations associated with large models and data, the community is increasingly investing in specialized hardware for model training. However, specialized hardware is expensive and hard to generalize to a multitude of tasks. The progress on the algorithmic front has failed to demonstrate a direct advantage over powerful hardware such as NVIDIA-V100 GPUs. This paper provides an exception. We propose SLIDE (Sub-LInear Deep learning Engine) that uniquely blends smart randomized algorithms, with multi-core parallelism and workload optimization. Using just a CPU, SLIDE drastically reduces the computations during both training and inference outperforming an optimized implementation of Tensorflow (TF) on the best available GPU. Our evaluations on industry-scale recommendation datasets, with large fully connected architectures, show that training with SLIDE on a 44 core CPU is more than 3.5 times (1 hour vs. 3.5 hours) faster than the same network trained using TF on Tesla V100 at any given accuracy level. On the same CPU hardware, SLIDE is over 10x faster than TF. We provide codes and scripts for reproducibility.

cs.DC↗

arXiv2019

Lsh-sampling Breaks the Computation Chicken-and-egg Loop in Adaptive Stochastic Gradient Estimation

Stochastic Gradient Descent or SGD is the most popular optimization algorithm for large-scale problems. SGD estimates the gradient by uniform sampling with sample size one. There have been several other works that suggest faster epoch-wise convergence by using weighted non-uniform sampling for better gradient estimates. Unfortunately, the per-iteration cost of maintaining this adaptive distribution for gradient estimation is more than calculating the full gradient itself, which we call the chicken-and-the-egg loop. As a result, the false impression of faster convergence in iterations, in reality, leads to slower convergence in time. In this paper, we break this barrier by providing the first demonstration of a scheme, Locality sensitive hashing (LSH) sampled Stochastic Gradient Descent (LGD), which leads to superior gradient estimation while keeping the sampling cost per iteration similar to that of the uniform sampling. Such an algorithm is possible due to the sampling view of LSH, which came to light recently. As a consequence of superior and fast estimation, we reduce the running time of all existing gradient descent algorithms, that relies on gradient estimates including Adam, Ada-grad, etc. We demonstrate the effectiveness of our proposal with experiments on linear models as well as the non-linear BERT, which is a recent popular deep learning based language representation model.

cs.LG↗

arXiv2019

Sub-Linear Privacy-Preserving Near-Neighbor Search

In Near-Neighbor Search (NNS), a new client queries a database (held by a server) for the most similar data (near-neighbors) given a certain similarity metric. The Privacy-Preserving variant (PP-NNS) requires that neither server nor the client shall learn information about the other party's data except what can be inferred from the outcome of NNS. The overwhelming growth in the size of current datasets and the lack of a truly secure server in the online world render the existing solutions impractical; either due to their high computational requirements or non-realistic assumptions which potentially compromise privacy. PP-NNS having query time {\it sub-linear} in the size of the database has been suggested as an open research direction by Li et al. (CCSW'15). In this paper, we provide the first such algorithm, called Secure Locality Sensitive Indexing (SLSI) which has a sub-linear query time and the ability to handle honest-but-curious parties. At the heart of our proposal lies a secure binary embedding scheme generated from a novel probabilistic transformation over locality sensitive hashing family. We provide information theoretic bound for the privacy guarantees and support our theoretical claims using substantial empirical evidence on real-world datasets.

cs.CR↗

arXiv2017

Unique Entity Estimation with Application to the Syrian Conflict

Entity resolution identifies and removes duplicate entities in large, noisy databases and has grown in both usage and new developments as a result of increased data availability. Nevertheless, entity resolution has tradeoffs regarding assumptions of the data generation process, error rates, and computational scalability that make it a difficult task for real applications. In this paper, we focus on a related problem of unique entity estimation, which is the task of estimating the unique number of entities and associated standard errors in a data set with duplicate entities. Unique entity estimation shares many fundamental challenges of entity resolution, namely, that the computational cost of all-to-all entity comparisons is intractable for large databases. To circumvent this computational barrier, we propose an efficient (near-linear time) estimation algorithm based on locality sensitive hashing. Our estimator, under realistic assumptions, is unbiased and has provably low variance compared to existing random sampling based approaches. In addition, we empirically show its superiority over the state-of-the-art estimators on three real applications. The motivation for our work is to derive an accurate estimate of the documented, identifiable deaths in the ongoing Syrian conflict. Our methodology, when applied to the Syrian data set, provides an estimate of $191,874 \pm 1772$ documented, identifiable deaths, which is very close to the Human Rights Data Analysis Group (HRDAG) estimate of 191,369. Our work provides an example of challenges and efforts involved in solving a real, noisy challenging problem where modeling assumptions may not hold.

stat.AP↗

arXiv2016

Revisiting Winner Take All (WTA) Hashing for Sparse Datasets

WTA (Winner Take All) hashing has been successfully applied in many large scale vision applications. This hashing scheme was tailored to take advantage of the comparative reasoning (or order based information), which showed significant accuracy improvements. In this paper, we identify a subtle issue with WTA, which grows with the sparsity of the datasets. This issue limits the discriminative power of WTA. We then propose a solution for this problem based on the idea of Densification which provably fixes the issue. Our experiments show that Densified WTA Hashing outperforms Vanilla WTA both in image classification and retrieval tasks consistently and significantly.

cs.CV↗