Tie Liu - NAOC, PKU

Tie Liu
Are you Tie Liu?

Claim your profile, edit publications, add additional information:

Contact Details

Name
Tie Liu
Affiliation
NAOC, PKU
Location

Pubs By Year

External Links

Pub Categories

 
Computer Science - Learning (19)
 
Astrophysics of Galaxies (16)
 
Solar and Stellar Astrophysics (16)
 
Computer Science - Computation and Language (8)
 
Computer Science - Information Theory (6)
 
Mathematics - Information Theory (6)
 
Computer Science - Information Retrieval (4)
 
Statistics - Machine Learning (4)
 
Computer Science - Artificial Intelligence (3)
 
Computer Science - Distributed; Parallel; and Cluster Computing (3)
 
Computer Science - Computer Science and Game Theory (3)
 
Computer Science - Neural and Evolutionary Computing (2)
 
Physics - Statistical Mechanics (1)
 
Mathematics - Optimization and Control (1)
 
Quantitative Biology - Genomics (1)
 
Quantitative Biology - Neurons and Cognition (1)

Publications Authored By Tie Liu

Stellar feedback from high-mass stars can strongly influence the surrounding interstellar medium and regulate star formation. Our new ALMA observations reveal sequential high-mass star formation taking place within one sub-virial filamentary clump (the G9.62 clump) in the G9. Read More

2017Apr
Authors: Derek Ward-Thompson, Kate Pattle, Pierre Bastien, Ray S. Furuya, Woojin Kwon, Shih-Ping Lai, Keping Qiu, David Berry, Minho Choi, Simon Coudé, James Di Francesco, Thiem Hoang, Erica Franzmann, Per Friberg, Sarah F. Graves, Jane S. Greaves, Martin Houde, Doug Johnstone, Jason M. Kirk, Patrick M. Koch, Jungmi Kwon, Chang Won Lee, Di Li, Brenda C. Matthews, Joseph C. Mottram, Harriet Parsons, Andy Pon, Ramprasad Rao, Mark Rawlings, Hiroko Shinnaga, Sarah Sadavoy, Sven van Loo, Yusuke Aso, Do-Young Byun, Eswariah Chakali, Huei-Ru Chen, Mike C. -Y. Chen, Wen Ping Chen, Tao-Chung Ching, Jungyeon Cho, Antonio Chrysostomou, Eun Jung Chung, Yasuo Doi, Emily Drabek-Maunder, Stewart P. S. Eyres, Jason Fiege, Rachel K. Friesen, Gary Fuller, Tim Gledhill, Matt J. Griffin, Qilao Gu, Tetsuo Hasegawa, Jennifer Hatchell, Saeko S. Hayashi, Wayne Holland, Tsuyoshi Inoue, Shu-ichiro Inutsuka, Kazunari Iwasaki, Il-Gyo Jeong, Ji-hyun Kang, Miju Kang, Sung-ju Kang, Koji S. Kawabata, Francisca Kemper, Gwanjeong Kim, Jongsoo Kim, Kee-Tae Kim, Kyoung Hee Kim, Mi-Ryang Kim, Shinyoung Kim, Kevin M. Lacaille, Jeong-Eun Lee, Sang-Sung Lee, Dalei Li, Hua-bai Li, Hong-Li Liu, Junhao Liu, Sheng-Yuan Liu, Tie Liu, A-Ran Lyo, Steve Mairs, Masafumi Matsumura, Gerald H. Moriarty-Schieven, Fumitaka Nakamura, Hiroyuki Nakanishi, Nagayoshi Ohashi, Takashi Onaka, Nicolas Peretto, Tae-Soo Pyo, Lei Qian, Brendan Retter, John Richer, Andrew Rigby, Jean-François Robitaille, Giorgio Savini, Anna M. M. Scaife, Archana Soam, Motohide Tamura, Ya-Wen Tang, Kohji Tomisaka, Hongchi Wang, Jia-Wei Wang, Anthony P. Whitworth, Hsi-Wei Yen, Hyunju Yoo, Jinghua Yuan, Chuan-Peng Zhang, Guoyin Zhang, Jianjun Zhou, Lei Zhu, Philippe André, C. Darren Dowell, Sam Falle, Yusuke Tsukamoto

We present the first results from the B-fields In STar-forming Region Observations (BISTRO) survey, using the Sub-millimetre Common-User Bolometer Array 2 (SCUBA-2) camera, with its associated polarimeter (POL-2), on the James Clerk Maxwell Telescope (JCMT) in Hawaii. We discuss the survey's aims and objectives. We describe the rationale behind the survey, and the questions which the survey will aim to answer. Read More

In this paper, we study a new learning paradigm for Neural Machine Translation (NMT). Instead of maximizing the likelihood of the human translation as in previous works, we minimize the distinction between human translation and the translation given by a NMT model. To achieve this goal, inspired by the recent success of generative adversarial networks (GANs), we employ an adversarial training architecture and name it as Adversarial-NMT. Read More

Machine learning is essentially the sciences of playing with data. An adaptive data selection strategy, enabling to dynamically choose different data at various training stages, can reach a more effective model in a more efficient way. In this paper, we propose a deep reinforcement learning framework, which we call \emph{\textbf{N}eural \textbf{D}ata \textbf{F}ilter} (\textbf{NDF}), to explore automatic and adaptive data selection in the training process. Read More

We consider the $(n,k,d,\ell)$ secure exact-repair regenerating code problem, which generalizes the $(n,k,d)$ exact-repair regenerating code problem with the additional constraint that the stored file needs to be kept information-theoretically secure against an eavesdropper, who can access the data transmitted to regenerate a total of $\ell$ different failed nodes. For all known results on this problem, the achievable tradeoff regions between the normalized storage capacity and repair bandwidth have a single corner point, achieved by a scheme proposed by Shah, Rashmi and Kumar (the SRK point). Since the achievable tradeoff regions of the exact-repair regenerating code problem without any secrecy constraints are known to have multiple corner points in general, these existing results suggest a phase-change-like behavior, i. Read More

The proximal gradient algorithm has been popularly used for convex optimization. Recently, it has also been extended for nonconvex problems, and the current state-of-the-art is the nonmonotone accelerated proximal gradient algorithm. However, it typically requires two exact proximal steps in each iteration, and can be inefficient when the proximal step is expensive. Read More

We observed thirteen Planck cold clumps with the James Clerk Maxwell Telescope/SCUBA-2 and with the Nobeyama 45 m radio telescope. The N$_2$H$^+$ distribution obtained with the Nobeyama telescope is quite similar to SCUBA-2 dust distribution. The 82 GHz HC$_3$N, 82 GHz CCS, and 94 GHz CCS emission are often distributed differently with respect to the N$_2$H$^+$ emission. Read More

Selling reserved instances (or virtual machines) is a basic service in cloud computing. In this paper, we consider a more flexible pricing model for instance reservation, in which a customer can propose the time length and number of resources of her request, while in today's industry, customers can only choose from several predefined reservation packages. Under this model, we design randomized mechanisms for customers coming online to optimize social welfare and providers' revenue. Read More

Decision tree (and its extensions such as Gradient Boosting Decision Trees and Random Forest) is a widely used machine learning algorithm, due to its practical effectiveness and model interpretability. With the emergence of big data, there is an increasing need to parallelize the training process of decision tree. However, most existing attempts along this line suffer from high communication costs. Read More

While neural machine translation (NMT) is making good progress in the past two years, tens of millions of bilingual sentence pairs are needed for its training. However, human labeling is very costly. To tackle this training data bottleneck, we develop a dual-learning mechanism, which can enable an NMT system to automatically learn from unlabeled data through a dual-learning game. Read More

Recurrent neural networks (RNNs) have achieved state-of-the-art performances in many natural language processing tasks, such as language modeling and machine translation. However, when the vocabulary is large, the RNN model will become very big (e.g. Read More

The feature-selection problem is formulated from an information-theoretic perspective. We show that the problem can be efficiently solved by an extension of the recently proposed info-clustering paradigm. This reveals the fundamental duality between feature selection and data clustering,which is a consequence of the more general duality between the principal partition and the principal lattice of partitions in combinatorial optimization. Read More

With the fast development of deep learning, people have started to train very big neural networks using massive data. Asynchronous Stochastic Gradient Descent (ASGD) is widely used to fulfill this task, which, however, is known to suffer from the problem of delayed gradient. That is, when a local worker adds the gradient it calculates to the global model, the global model may have been updated by other workers and this gradient becomes "delayed". Read More

Many machine learning tasks can be formulated as Regularized Empirical Risk Minimization (R-ERM), and solved by optimization algorithms such as gradient descent (GD), stochastic gradient descent (SGD), and stochastic variance reduction (SVRG). Conventional analysis on these optimization algorithms focuses on their convergence rates during the training process, however, people in the machine learning community may care more about the generalization performance of the learned model on unseen test data. In this paper, we investigate on this issue, by using stability as a tool. Read More

Regularized empirical risk minimization (R-ERM) is an important branch of machine learning, since it constrains the capacity of the hypothesis space and guarantees the generalization ability of the learning algorithm. Two classic proximal optimization algorithms, i.e. Read More

In recent year, parallel implementations have been used to speed up the training of deep neural networks (DNN). Typically, the parameters of the local models are periodically communicated and averaged to get a global model until the training curve converges (denoted as MA-DNN). However, since DNN is a highly non-convex model, the global model obtained by averaging parameters does not have guarantee on its performance improvement over the local models and might even be worse than the average performance of the local models, which leads to the slow-down of convergence and the decrease of the final performance. Read More

We observed 146 Galactic clumps in HCN (4-3) and CS (7-6) with the Atacama Submillimeter Telescope Experiment (ASTE) 10-m telescope. A tight linear relationship between star formation rate and gas mass traced by dust continuum emission was found for both Galactic clumps and the high redshift (z>1) star forming galaxies (SFGs), indicating a constant gas depletion time of ~100 Myr for molecular gas in both Galactic clumps and high z SFGs. However, low z galaxies do not follow this relation and seem to have a longer global gas depletion time. Read More

We formulate an info-clustering paradigm based on a multivariate information measure, called multivariate mutual information, that naturally extends Shannon's mutual information between two random variables to the multivariate case involving more than two random variables. With proper model reductions, we show that the paradigm can be applied to study the human genome and connectome in a more meaningful way than the conventional algorithmic approach. Not only can info-clustering provide justifications and refinements to some existing techniques, but it also inspires new computationally feasible solutions. Read More

Most bipolar outflows are associated with individual young stellar objects and have small opening angles. Here we report the discovery of an extremely wide-angle ($\sim$180$\arcdeg$) bipolar outflow ("EWBO") in a cluster forming region AFGL 5142 from low-velocity emission of the HCN (3-2) and HCO$^{+}$ (3-2) lines. This bipolar outflow is along a north-west to south-east direction with a line-of-sight flow velocity of about 3 km~s$^{-1}$ and is spatially connected to the high-velocity jet-like outflows. Read More

We study the effects of an asymmetric radiation field on the properties of a molecular cloud envelope. We employ observations of carbon monoxide (12CO and 13CO), atomic carbon, ionized carbon, and atomic hydrogen to analyze the chemical and physical properties of the core and envelope of L1599B, a molecular cloud forming a portion of the ring at approximately 27 pc from the star Lambda Ori. The O III star provides an asymmetric radiation field that produces a moderate enhancement of the external radiation field. Read More

We propose Sentence Level Recurrent Topic Model (SLRTM), a new topic model that assumes the generation of each word within a sentence to depend on both the topic of the sentence and the whole history of its preceding words in the sentence. Different from conventional topic models that largely ignore the sequential order of words or their topic coherence, SLRTM gives full characterization to them by using a Recurrent Neural Networks (RNN) based framework. Experimental results have shown that SLRTM outperforms several strong baselines on various tasks. Read More

An important problem in the field of bioinformatics is to identify interactive effects among profiled variables for outcome prediction. In this paper, a logistic regression model with pairwise interactions among a set of binary covariates is considered. Modeling the structure of the interactions by a graph, our goal is to recover the interaction graph from independently identically distributed (i. Read More

We present the first survey of dense gas towards Planck Galactic Cold Clumps (PGCCs). Observations in the J=1-0 transitions of HCO+ and HCN towards 621 molecular cores associated with PGCCs were performed using the Purple Mountain Observatory 13.7-m telescope. Read More

The physical mechanisms that induce the transformation of a certain mass of gas in new stars are far from being well understood. Infrared bubbles associated with HII regions have been considered to be good samples of investigating triggered star formation. In this paper we report on the investigation of the dust properties of the infrared bubble N4 around the HII region G11. Read More

The problem of multilevel diversity coding with regeneration is considered in this work. Two new outer bounds on the optimal tradeoffs between the normalized storage capacity and repair bandwidth are established, by which the optimality of separate coding at the minimum-bandwidth-regeneration (MBR) point follows immediately. This resolves a question left open in a previous work by Tian and Liu. Read More

We present Submillimeter Array (SMA) observations of the HCN\,(3--2) and HCO$^+$\,(3--2) molecular lines toward the W3(H$_2$O) and W3(OH) star-forming complexes. Infall and outflow motions in the W3(H$_2$O) have been characterized by observing HCN and HCO$^+$ transitions. High-velocity blue/red-shifted emission, tracing the outflow, show multiple knots, which might originate in episodic and precessing outflows. Read More

We propose a coalition game model for the problem of communication for omniscience (CO). In this game model, the core contains all achievable rate vectors for CO with sum-rate being equal to a given value. Any rate vector in the core distributes the sum-rate among users in a way that makes all users willing to cooperate in CO. Read More

We present Submillimeter Array (SMA) molecular line observations in two 2 GHz-wide bands centered at 217.5 and 227.5 GHz, toward the massive star forming region W51 North. Read More

CIT 6 is a carbon star in the transitional phase from the asymptotic giant branch (AGB) to the protoplanetary nebulae (pPN). Observational evidences of two point sources in the optical, circumstellar arc segments in an HC$_3$N line emission, and a bipolar nebula in near-infrared provide strong support for the presence of a binary companion. Hence, CIT 6 is very attractive for studying the role of companions in the AGB-pPN transition. Read More

Based on the probability distribution observed in complex systems and an assumption that the probability distributions of complex systems satisfy a new generalized multiplication, it is proved that the statistical theory of complex systems can be established in the analogous extensive framework. Read More

People believe that depth plays an important role in success of deep neural networks (DNN). However, this belief lacks solid theoretical justifications as far as we know. We investigate role of depth from perspective of margin bound. Read More

Intelligence Quotient (IQ) Test is a set of standardized questions designed to evaluate human intelligence. Verbal comprehension questions appear very frequently in IQ tests, which measure human's verbal ability including the understanding of the words with multiple senses, the synonyms and antonyms, and the analogies among words. In this work, we explore whether such tests can be solved automatically by artificial intelligence technologies, especially the deep learning technologies that are recently developed and successfully applied in a number of fields. Read More

Word embedding, which refers to low-dimensional dense vector representations of natural words, has demonstrated its power in many natural language processing tasks. However, it may suffer from the inaccurate and incomplete information contained in the free text corpus as training data. To tackle this challenge, there have been quite a few works that leverage knowledge graphs as an additional information source to improve the quality of word embedding. Read More

Thompson sampling is one of the earliest randomized algorithms for multi-armed bandits (MAB). In this paper, we extend the Thompson sampling to Budgeted MAB, where there is random cost for pulling an arm and the total cost is constrained by a budget. We start with the case of Bernoulli bandits, in which the random rewards (costs) of an arm are independently sampled from a Bernoulli distribution. Read More

Digital contents in large scale distributed storage systems may have different reliability and access delay requirements, and for this reason, erasure codes with different strengths need to be utilized to achieve the best storage efficiency. At the same time, in such large scale distributed storage systems, nodes fail on a regular basis, and the contents stored on them need to be regenerated and stored on other healthy nodes, the efficiency of which is an important factor affecting the overall quality of service. In this work, we formulate the problem of multilevel diversity coding with regeneration to address these considerations, for which the storage vs. Read More

We report on the Submillimeter Array (SMA) observations of molecular lines at 270 GHz toward W3(OH) and W3(H$_2$O) complex. Although previous observations already resolved the W3(H$_2$O) into two or three sub-components, the physical and chemical properties of the two sources are not well constrained. Our SMA observations clearly resolved W3(OH) and W3(H$_2$O) continuum cores. Read More

When building large-scale machine learning (ML) programs, such as big topic models or deep neural nets, one usually assumes such tasks can only be attempted with industrial-sized clusters with thousands of nodes, which are out of reach for most practitioners or academic researchers. We consider this challenge in the context of topic modeling on web-scale corpora, and show that with a modest cluster of as few as 8 machines, we can train a topic model with 1 million topics and a 1-million-word vocabulary (for a total of 1 trillion parameters), on a document collection with 200 billion tokens -- a scale not yet reported even with thousands of machines. Our major contributions include: 1) a new, highly efficient O(1) Metropolis-Hastings sampling algorithm, whose running cost is (surprisingly) agnostic of model size, and empirically converges nearly an order of magnitude faster than current state-of-the-art Gibbs samplers; 2) a structure-aware model-parallel scheme, which leverages dependencies within the topic model, yielding a sampling strategy that is frugal on machine memory and network communication; 3) a differential data-structure for model storage, which uses separate data structures for high- and low-frequency words to allow extremely large models to fit in memory, while maintaining high inference speed; and 4) a bounded asynchronous data-parallel scheme, which allows efficient distributed processing of massive data via a parameter server. Read More

How high-mass stars form remains unclear currently. Calculation suggests that the radiation pressure of a forming star can halt spherical infall, preventing its further growth when it reaches 10 M$_{\odot}$. Two major theoretical models on the further growth of stellar mass were proposed. Read More

The physical and chemical properties of prestellar cores, especially massive ones, are still far from being well understood due to the lack of a large sample. The low dust temperature ($<$14 K) of Planck cold clumps makes them promising candidates for prestellar objects or for sources at the very initial stages of protostellar collapse. We have been conducting a series of observations toward Planck cold clumps (PCCs) with ground-based radio telescopes. Read More

For Internet applications like sponsored search, cautions need to be taken when using machine learning to optimize their mechanisms (e.g., auction) since self-interested agents in these applications may change their behaviors (and thus the data distribution) in response to the mechanisms. Read More

This work reports a high spatial resolution observations toward Orion KL region with high critical density lines of CH$_{3}$CN (12$_{4}$-11$_{4}$) and CH$_{3}$OH (8$_{-1, 8}$-7$_{0, 7}$) as well as continuum at $\sim$1.3 mm band. The observations were made using the Atacama Large Millimeter/Submillimeter Array with a spatial resolution of $\sim$1. Read More

WordRep is a benchmark collection for the research on learning distributed word representations (or word embeddings), released by Microsoft Research. In this paper, we describe the details of the WordRep collection and show how to use it in different types of machine learning research related to word embedding. Specifically, we describe how the evaluation tasks in WordRep are selected, how the data are sampled, and how the evaluation tool is built. Read More

Neural network techniques are widely applied to obtain high-quality distributed representations of words, i.e., word embeddings, to address text mining, information retrieval, and natural language processing tasks. Read More

Sponsored search is an important monetization channel for search engines, in which an auction mechanism is used to select the ads shown to users and determine the prices charged from advertisers. There have been several pieces of work in the literature that investigate how to design an auction mechanism in order to optimize the revenue of the search engine. However, due to some unrealistic assumptions used, the practical values of these studies are not very clear. Read More

2014Apr

Context: We investigated the molecular gas associated with 6.7 GHz methanol masers throughout the Galaxy using a J=1-0 transition of the CO isotopologues. Methods:Using the 13. Read More

Click prediction is one of the fundamental problems in sponsored search. Most of existing studies took advantage of machine learning approaches to predict ad click for each event of ad view independently. However, as observed in the real-world sponsored search system, user's behaviors on ads yield high dependency on how the user behaved along with the past time, especially in terms of what queries she submitted, what ads she clicked or ignored, and how long she spent on the landing pages of clicked ads, etc. Read More