Jia-Yu (Tim) Pan
Research Areas
Authored Publications
Sort By
Preview abstract
A visualization sequence is an effective representation of meaningful data stories. Existing visualization sequencing approaches use heuristics to arrange charts in a meaningful order. While they perform well in specific scenarios, they do not customize the generated sequences to individual users’ preferences. In this work, we present VisGuide , an assistive data exploration system that helps a user create contextual visualization sequence trees by sequentially recommending meaningful charts tailoring to the user’s preference on data exploration. Our results show that VisGuide can recommend chart sequences that interest users and are also considered meaningful by domain experts.
View details
Complement Objective Training
Hao-Yun Chen
Pei-Hsin Wang
Chun-Hao Liu
Shih-Chieh Chang
Yu-Ting Chen
Wei Wei
International Conference on Learning Representations (ICLR) (2019)
Preview abstract
Learning with a primary objective, such as softmax cross entropy for classification and sequence generation, has been the norm for training deep neural networks for years. Although being a widely-adopted approach, using cross entropy as the primary objective exploits mostly the information from the ground-truth class for maximizing data likelihood, and largely ignores information from the complement (incorrect) classes. We argue that, in addition to the primary objective, training also using a complement objective that leverages information from the complement classes can be effective in improving model performance. This motivates us to study a new training paradigm that maximizes the likelihood of the ground-truth class while neutralizing the probabilities of the complement classes. We conduct extensive experiments on multiple tasks ranging from computer vision to natural language processing. The experimental results confirm that, compared to the conventional training with just one primary objective, training also with the complement objective further improves the performance of the state-of-the-art models across all tasks. In addition to the accuracy improvement, we also show that models trained with both primary and complement objectives are more robust
to adversarial attacks.
View details
Improving Adversarial Robustness via Guided Complement Entropy
Hao-Yun Chen
Jhao-Hong Liang
Shih-Chieh Chang
Yu-Ting Chen
Wei Wei
International Conference on Computer Vision (ICCV) (2019)
Preview abstract
Adversarial robustness has emerged as an important topic in deep learning as carefully crafted attack samples can significantly disturb the performance of a model. Many recent methods have proposed to improve adversarial robustness by utilizing adversarial training or model distillation, which adds additional procedures to model training. In this paper, we propose a new training paradigm called Guided Complement Entropy (GCE) that is capable of achieving “adversarial defense for free,” which involves no additional procedures in the process of improving adversarial robustness. In addition to maximizing model probabilities on the ground-truth class like cross entropy, we neutralize its probabilities on the incorrect classes along with a “guided” term to balance between these two terms. We show in the experiments that our method achieves better model robustness with even better performance compared to the commonly used cross entropy training objective. We also show that our method can be used orthogonal to adversarial training across well known methods with noticeable robustness gain. To the best of our knowledge, our approach is the first one that improves model robustness without compromising performance.
View details
MONAS: Multi-Objective Neural Architecture Search using Reinforcement Learning
Chi-Hung Hsu
Shih-Chieh Chang
Jhao-Hong Liang
Hsin-Ping Chou
Chun-Hao Liu
Shu-Huan Chang
Yu-Ting Chen
Wei Wei
(2018)
Preview abstract
Recent studies on neural architecture search have shown that automatically designed neural networks perform as good as expert-crafted architectures. While most existing works aim at finding architectures that optimize the prediction accuracy, these architectures may have complexity and is therefore not suitable being deployed on certain computing environment (e.g., with limited power budgets). We propose MONAS, a framework for Multi-Objective Neural Architectural Search that employs reward functions considering both prediction accuracy and other important objectives (e.g., power consumption) when searching for neural network architectures. Experimental results showed that, compared to the state-of-the-arts, models found by MONAS achieve comparable or better classification accuracy on computer vision applications, while satisfying the additional objectives such as peak power.
View details
Scalable Community Discovery from Multi-Faceted Graphs
Ahmed Metwally
Minh Doan
Christos Faloutsos
2015 IEEE International Conference on Big Data, IEEE, 445 Hoes Lane Piscataway, NJ 08854-4141 USA (to appear)
Preview abstract
A multi-faceted graph defines several facets on a set of nodes. Each facet is a set of edges that represent the relationships between the nodes in a specific context. Mining multi-faceted graphs have several applications, including finding fraudster rings that launch advertising traffic fraud attacks, tracking IP addresses of botnets over time, analyzing interactions on social networks and co-authorship of scientific papers. We propose NeSim, a distributed efficient clustering algorithm that does soft clustering on individual facets. We also propose optimizations to further improve the scalability, the efficiency and the clusters quality. We employ general purpose graph-clustering algorithms in a novel way to discover communities across facets. Due to the qualities of NeSim, we employ it as a backbone in the distributed MuFace algorithm, which discovers multi-faceted communities. We evaluate the proposed algorithms on several real and synthetic datasets, where NeSim is shown to be superior to MCL, JP and AP, the well-established clustering algorithms. We also report the success stories of MuFace in finding advertisement click rings.
View details
TSum: Fast, Principled Table Summarization.
Jieying Chen
Christos Faloutsos
Spiros Papadimitriou
Proceedings of the Seventh International Workshop on Data Mining for Online Advertising, ACM (2013)
Preview abstract
Given a table where rows correspond to records and columns correspond to attributes, we want to find a small number of patterns that succinctly summarize the dataset. For example, given a set of patient records with several attributes each, how can we find (a) that the "most representative" pattern is, say, (male, adult, *), followed by (*, child, low-cholesterol), etc.? We propose TSum, a method that provides a sequence of patterns ordered by their "representativeness." It can decide both which these patterns are, as well as how many are necessary to properly summarize the data. Our main contribution is formulating a general framework, TSum, using compression principles. TSum can easily accommodate different optimization strategies for selecting and refining patterns. The discovered patterns can be used to both represent the data efficiently, as well as interpret it quickly. Extensive experiments demonstrate the effectiveness and intuitiveness of our discovered patterns.
View details
The Goals and Challenges of Click Fraud Penetration Testing Systems
Carmelo Kintana
David Turner
Ahmed Metwally
Neil Daswani
Erika Chin
Andrew Bortz
International Symposium on Software Reliability Engineering, International Symposium on Software Reliability Engineering (2009)
Preview abstract
It is important for search and pay-per-click engines to penetration test their click fraud detection systems, in order to find potential vulnerabilities and correct them before fraudsters can exploit them. In this paper, we describe: (1) some goals and desirable qualities of a click fraud penetration testing system, based on our experience, and (2) our experiences with the challenges of building and using a click fraud penetration testing system called Camelot that has been in use at Google.
View details
