Kevin D. Seppi

• PhD
• Professor (Associate) at Brigham Young University

Topic modeling is a common tool for understanding large bodies of text, but is typically provided as a “take it or leave it” proposition. Incorporating human knowledge in unsupervised learning is a promising approach to create high-quality topic models. Existing interactive systems and modeling algorithms support a wide range of refinement operatio...

Corpus labeling projects frequently use low-cost workers from microtask marketplaces; however, these workers are often inexperienced or have misaligned incentives. Crowdsourcing models must be robust to the resulting systematic and non-systematic inaccuracies. We introduce a novel crowdsourcing model that adapts the discrete supervised topic model...

In modern practice, labeling a dataset often involves aggregating annotator judgments obtained from crowdsourcing. State-of-the-art aggregation is performed via inference on probabilistic models, some of which are data-aware, meaning that they leverage features of the data (e.g., words in a document) in addition to annotator judgments. Previous wor...

Return-on-Investment (ROI) is a cost-conscious approach to active learning (AL) that considers both estimates of cost and of benefit in active sample selection. We investigate the theoretical conditions for successful cost-conscious AL using ROI by examining the conditions under which ROI would optimize the area under the cost/benefit curve. We the...

Particle Swarm Optimization uses noisy historical information to select potentially optimal function samples. Though information-theoretic principles suggest that less noise indicates greater certainty, PSO's momentum term is usually both the least informed and the most deterministic. This dichotomy suggests that while momentum has a profound impac...

Particle Swarm Optimization is fundamentally a stochastic algorithm, where each particle takes into account noisy information from its own history as well as that of its neighborhood. Though basic information-theoretic principles would suggest that less noise indicates greater certainty, the momentum term is simultaneously the least directly-inform...

From multi-core processors to parallel GPUs to computing clusters, computing resources are increasingly parallel. These parallel resources are being used to address increasingly challenging applications. This presents an opportunity to design optimization algorithms that use parallel processors efficiently. In spite of the intuitively parallel natu...

Data annotation in modern practice often involves multiple, imperfect human annotators. Multiple annotations can be used to infer estimates of the ground-truth labels and to estimate individual annotator error characteristics (or reliability). We introduce MOMRESP, a model that improves upon item response models to incorporate information from both...

We describe an under-studied problem in language resource management: that of providing automatic assistance to annotators working in exploratory settings. When no satisfactory tagset already exists, such as in under-resourced or undocumented languages, it must be developed iteratively while annotating data. This process naturally gives rise to a s...

The task of corpus-dictionary linkage (CDL) is to annotate each word in a corpus with a link to an appropriate dictionary entry that documents the sense and usage of the word. Corpus-dictionary linked resources include concordances, dictionaries with word usage examples, and corpora annotated with lemmas or word senses. Such CDL resources are essen...

We introduce an intelligent cooperative control system for ground target tracking in a cluttered urban environment with a team of autonomous Unmanned Air Vehicles (UAVs). We extend the work of Yu et al. to use observations of target position to learn a model of target motion. Simulated cooperative control of a team of 9 UAVs in a 100-block city fil...

Machine assistance is vital to managing the cost of corpus annotation projects. Identifying effective forms of machine assistance through principled evaluation is particularly important and challenging in under-resourced domains and highly heterogeneous corpora, as the quality of machine assistance varies. We perform a fine-grained evaluation of tw...

We introduce an intelligent cooperative control system for ground target tracking in a cluttered urban environment with a team of Unmanned Air Vehicles (UAVs). We extend the work of Yu et. al. [1] to add a machine learning component that uses observations of target position to learn a model of target motion. Our learner is the Sequence Memoizer [2]...

Supervised topic models are promising tools for text analytics that simultaneously model topical patterns in document collections and relationships between those topics and document metadata, such as timestamps. We examine empirically the effect of OCR noise on the ability of supervised topic models to produce high quality output through a series o...

Despite popular use of Latent Dirichlet Allocation (LDA) for automatic discovery of latent topics in document corpora, such topics lack connections with relevant knowledge sources such as Wikipedia, and they can be difficult to interpret due to the lack of meaningful topic labels. Furthermore, the topic analysis suffers from a lack of identifiabili...

The MapReduce parallel programming model is designed for large-scale data processing, but its benefits, such as fault tolerance and automatic message routing, are also helpful for computationally-intensive algorithms. However, popular MapReduce frameworks such as Hadoop are slow for many scientific applications and are inconvenient on supercomputer...

In the natural world there are many swarms in any geographical region. In contrast, Particle Swarm Optimization (PSO) is usually used with a single swarm of particles. We define a simple new topology called Apiary and show that parallel communities of swarms give rise to emergent behavior that is fundamentally different from the behavior of a singl...

Manual annotation of large textual corpora can be cost-prohibitive, especially for rare and under-resourced languages. One potential solution is pre-annotation: asking human annotators to correct sentences that have already been annotated, usually by a machine. Another potential solution is correction propagation: using annotator corrections to dyn...

This work represents the first step towards a task li- brary system in the reinforcement learning domain. Task libraries could be useful in speeding up the learning of new tasks through task transfer. Related transfer can increase learning rate and can help pre- vent convergence to sub-optimal policies in reinforce- ment learning. Unrelated transfe...

Bayesian methods are theoretically optimal in many situations. Bayesian model averaging is generally con-sidered the standard model for creating ensembles of learners using Bayesian methods, but this technique is often out-performed by more ad hoc methods in empirical studies. The reason for this failure has important theoretical implications for o...

We formally define multi-capacity bin packing, a generalization of conventional bin packing, and develop an algo- rithm called Reordering Grouping Genetic Algorithm (RGGA) to assign VMs to servers. We first test RGGA on conventional bin packing problems and show that it yields excellent results but much more efficiently. We then generate a multi-co...

Particle swarm optimization (PSO) has previously been parallelized primarily by distributing the computation corresponding to particles across multiple processors. In these approaches, the only benefit of additional processors is an increased swarm size. However, in many cases this is not efficient when scaled to very large swarm sizes (on very lar...

Expert human input can contribute in various ways to facilitate automatic annotation of natural language text. For example, a part-of-speech tagger can be trained on labeled input provided offline by experts. In addition, expert input can be solicited by way of active learning to make the most of annotator expertise. However, hiring individuals to...

We introduce CCASH (Cost-Conscious Annotation Supervised by Humans), an extensible web application framework for cost-efficient annotation. CCASH provides a framework in which cost-efficient annotation methods such as Active Learning can be explored via user studies and afterwards applied to large annotation projects. CCASH's architecture is d escr...

Particle swarm optimization (PSO) has previously been parallelized only by adding more particles to the swarm or by parallelizing the evaluation of the objective function. However, some functions are more efficiently optimized with more iterations and fewer particles. Accordingly, we take inspiration from speculative execution performed in modern p...

Many have used the principles of statistics and Bayesian decision theory to model specific learning problems. It is less common to see models of the processes of learning in general. One exception is the model of the supervised learning process known as the “Extended Bayesian Formalism” or EBF. This model is descriptive, in that it can describe and...

Named entity recognition applied to scanned and OCRed historical documents can contribute to the discoverability of historical information. However, entity recognition from some historical corpora is much more difficult than from natively digital text because of the marked presence of word errors and absence of page layout information. How difficul...

We define a probabilistic morphological analyzer using a data-driven approach for Syriac in order to facilitate the creation of an annotated corpus. Syriac is an under-resourced Semitic language for which there are no available language tools such as morphological analyzers. We introduce novel probabilistic models for segmentation, dictionary linka...

We cast the assignment of virtual machines (VMs) to physical servers as a variant of the classic bin packing problem. We then develop a model of VM load that can be used to produce assignments of VMs to servers. Using this problem formulation and model, we evaluate heuristic solutions to this problem. We evaluate the performance of these solutions...

Polar nano-regions (PNR) in relaxor materials Pb(Zn1/3Nb2/3)O3 and Pb(Mg1/3Nb2/3)O3 are of pressing applied interest due to their influence on the remarkable piezoelectric properties of their solid solutions with ferroelectric PbTiO3. X-ray single-crystal diffuse-scattering techniques have recently been shown to provide qualitative insight into the...

We take a Bayesian approach to the issues of bias, meta bias, transfer, overfit, and No-Free-Lunch in the context of supervised learning. If we accept certain relationships be-tween the function class, on training set data, and off train-ing set data, then a graphical model can be created that represents the supervised learning problem. This graphi...

Particle Swarm Optimization (PSO) has typically been used with small swarms of about 50 particles. However, PSO is more efficiently parallelized with large swarms. We formally describe existing topologies and identify variations which are better suited to large swarms in both sequential and parallel computing environments. We examine the performanc...

This paper presents Bayesian edge inference (BEI), a single frame super resolution method explicitly grounded in Bayesian inference that addresses issues common to existing methods. Though the best give excellent results at modest magnification factors, they suffer from gradient stepping and boundary coherence problems by factors of 4x. Central to...

We present a statistical model of empirical optimization that admits the creation of algorithms with explicit and intuitively defined desiderata. Because No Free Lunch theorems dictate that no optimization algorithm can be considered more efficient than any other when considering all possible functions, the desired function class plays a prominent...

Fixed, limited budgets often constrain the amount of expert annotation that can go into the construction of annotated corpora. Estimating the cost of annotation is the first step toward using annotation resources wisely. We present here a study of the cost of annotation. This study includes the participation of annotators at various skill levels an...

Traditional Active Learning (AL) techniques assume that the annotation of each datum costs the same. This is not the case when anno- tating sequences; some sequences will take longer than others. We show that the AL tech- nique which performs best depends on how cost is measured. Applying an hourly cost model based on the results of an annotation u...

The performance of value iteration can be dramatically improved by eliminating redundant or useless updates to the value function – up-dates which have traditionally been included in complexity analyses. We present an alternative, but related, algorithm which discards the notion of an "iteration" altogether and automatically exploits the details of...

We use a graphical model of the supervised learning problem to explore the the-oretical effect of utility in the form of end use and sample cost on supervised learning, No-Free-Lunch, sample complexity, and active learning. There are two sources of utility that can affect the above problems: utility that comes from end use and utility that comes fr...

In optimization problems involving large amounts of data, such as web content, commercial transaction information, or bioinformatics data, individual function evaluations may take minutes or even hours. particle swarm optimization (PSO) must be parallelized for such functions. However, large-scale parallel programs must communicate efficiently, bal...

We recast the problem of unconstrained continuous evolutionary optimization as inference in a fixed graphical model. This approach allows us to address several pervasive issues in optimization, including the traditionally difficult problem of selecting an algorithm that is most appropriate for a given task. This is accomplished by placing a prior d...

In the construction of a part-of-speech annotated corpus, we are constrained by a fixed budget. A fully annotated corpus is required, but we can afford to label only a subset. We train a Maximum Entropy Markov Model tagger from a labeled subset and automatically tag the remainder. This paper addresses the question of where to focus our manual taggi...

In optimization problems involving large amounts of data, Particle Swarm Optimization (PSO) must be parallelized because individual function evaluations may take minutes or even hours. However, large-scale parallelization is difficult because programs must communicate efficiently, balance workloads and tolerate node failures. To address these issue...

This paper shows that the basic Hough transform is implicitly a Bayesian process-that it computes an unnormalized posterior distribution over the parameters of a single shape given feature points. The proof motivates a purely Bayesian approach to the problem of finding parameterized shapes in digital images. A proof-of-concept implementation that f...

We analyze the drawbacks to using ANNs in high assurance systems and propose a solution based upon a Bayesian approach with a specific network topology that can be solved in closed form. The Bayesian approach leads to better answers in the traditional sense, while also allowing us to quantify risk and deal with it in a reasonable manner. We demonst...

In the construction of a part-of-speech annotated corpus, we are constrained by a fixed budget. A fully annotated corpus is required, but we can afford to label only a subset. We train a Maximum Entropy Markov Model tagger from a labeled subset and automatically tag the remainder. This paper addresses the question of where to focus our manual taggi...

Phylogenetic analysis is a central tool in studies of comparative genomics. When a new region of DNA is isolated and sequenced, researchers are often forced to throw away months of computation on an existing phylogeny of homologous sequences in order to incorporate this new sequence. The previously constructed trees are often discarded, and the res...

This paper presents a meta-heuristic for use in finding errors in models of complex concurrent systems using explicit guided model checking. The meta-heuristic improves explicit guided model checking by applying the empirical Bayes method to revise heuristic estimates of the distance from a given state to an error state. Guided search using the rev...

Spatial Extension PSO (SEPSO) and Attractive-Repulsive PSO (ARPSO) are methods for artificial injection of diver- sity into particle swarm optimizers that are intended to en- courage converged swarms to engage in exploration. While simple to implement, effective when tuned correctly, and benefiting from intuitive appeal, SEPSO behavior can be impro...

Dynamic pricing is a real-time machine learning problem with scarce prior data and a concrete learning cost. While the Kalman Filter can be employed to track hidden demand parameters and extensions to it can facilitate exploration for faster learning, the exploratory nature of particle swarm optimization makes it a natural choice for the dynamic pr...

This paper shows how the notion of value of cooperation, a measure of the percentage of a Arm's profits due strictly to the cooperative effects among the goods it sells, can be used to analyze the relative economic advantage afforded by various organizational structures. The value of cooperation is computed from transactions data by solving a regre...

We present a homomorphous mapping that converts problems with linear equality constraints into fully unconstrained and lower-dimensional problems for optimization with PSO. This approach, in contrast with feasibility preservation methods, allows any unconstrained optimization algorithm to be applied to a problem with linear equality constraints, ma...

This paper demonstrates the utility of systems and control theory in the analysis of economic systems. Two applications demonstrate how the analysis of simple dynamic models sheds light on important practical problems. The first problem considers the design of a retail laboratory, where the small gain theorem enables the falsification of pricing po...

We describe a new algorithm designed to quickly and robustly solve general linear problems of the form Ax = b. We describe both serial and parallel versions of the algorithm, which can be considered a prioritized version of an Alternating Multiplicative Schwarz procedure. We also adopt a general view of alternating Multiplicative Schwarz procedures...

The performance of value and policy iteration can be dramatically improved by eliminating redundant or useless backups, and by backing up states in the right order. We study several methods designed to accelerate these iterative solvers, including prioritization, partitioning, and variable reordering. We generate a family of algorithms by combining...

We discuss testing methods for exposing origin-seeking bias in PSO motion algorithms. The strategy of resizing the initialization space, proposed by Gehlhaar and Fogel and made popular in the PSO context by Angeline, is shown to be insufficiently general for revealing an algorithm's tendency to focus its efforts on regions at or near the origin. An...

We explore the use of information models as a guide for the development of single objective optimization algorithms, giving particular attention to the use of Bayesian models in a PSO context. The use of an explicit information model as the basis for particle motion provides tools for designing successful algorithms. One such algorithm is developed...

Recent research in task transfer and task clustering has necessitated the need for task similarity measures in reinforcement learning. Determining task similarity is necessary for selective transfer where only information from relevant tasks and portions of a task are transferred. Which task similarity measure to use is not immediately obvious. It...

Particle Swarm Optimization is gaining momentum as a simple and effective optimization technique. We present a new approach to PSO that significantly reduces the number of iterations required to reach good solutions. In contrast with much recent research, the focus of this work is on fundamental particle motion, making use of the Kalman Filter to u...

This work presents a formal verification algorithm for finding errors in models of complex concurrent systems. The algorithm improves explicit guided model checking by applying the empirical Bayes method to revise heuristic estimates of the distance from a given state to an error state. Guided search using the revised estimates finds errors with le...

Particle Swarm Optimization is gaining momentum as a simple and effective optimization technique. We present a new approach to PSO that significantly reduces the number of iterations required to reach good solutions. In contrast with much recent research, the focus of this work is on fundamental particle motion, making use of the Kalman Filter to u...

Not Available

Not Available

We present an examination of the state-of-the-art for using value iteration to solve large-scale discrete Markov Decision Processes. We introduce an architecture which combines three independent performance enhancements (the intelligent prioritization of computation, state partitioning, and massively parallel processing) into a single algorithm. We...

As algorithms scale to solve larger and larger MDPs, it be- comes impossible to store all of the model information of the MDP and the supporting data structures of the algorithm in RAM. This motivates the study of the disk-based-cache ef- ficiency of solution algorithms. We contrast the cache effi- ciency of normal value iteration with that of the...

The focus of this paper is the application of Bayesian concepts to database query optimization. In relational database systems users retrieve data by describing the desired data. The description of the desired data takes the form of statements which specify operations on “relations” (as defined by set theory). In a relational database system a “que...

Named entity recognition from scanned and OCRed historical documents can contribute to historical research. However, entity recogni-tion from historical documents is more diffi-cult than from natively digital data because of the presence of word errors and the absence of complete formatting information. We ap-ply four extraction algorithms to vario...

This report explores the challenges and opportu- nities inherent in developing a retail laboratory. In particular, retail is considered as a feedback phenomenon, and some fundamental limits on learning from data are identified. The small gain theorem is then reviewed, and the unique value of a laboratory enabling controlled, live experimentation is...

The Kalman Swarm (KSwarm) is a new approach to particle motion in PSO that reduces the number of iterations required to reach good solutions (1). Unfortunately, it has much higher computational com- plexity than basic PSO. This paper addresses the runtime of KSwarm in a new algorithm called \Linear Kalman Swarm" (LinkSwarm) which has linear complex...

We describe a new algorithm designed to quickly and robustly solve general linear problems of the form Ax = b. We describe both serial and parallel versions of the algorithm, which can be considered a prioritized version of an Alternating Multiplicative Schwarz procedure. We also adopt a general view of alternating Multiplicative Schwarz procedures...

This paper proposes a measure to quantify the value of cooperation experienced by a firm. Using a "reverse" merger simulation approach, the percentage of a firm's profits due to cooperation can be precisely determined. This is accomplished by considering the profit maximizing dynamics of firms in the market as defining a value function of a coaliti...

Topic models have been shown to reveal the semantic content in large corpora. Many individualized visualizations of topic models have been reported in the lit-erature, showing the potential of topic models to give valuable insight into a cor-pus. However, good, general tools for browsing the entire output of a topic model along with the analyzed co...