Discriminative training of the hidden vector state model for semantic parsing

In this paper, we discuss how discriminative training can be applied to the hidden vector state (HVS) model in different task domains. The HVS model is a discrete hidden Markov model (HMM) in which each HMM state represents the state of a push-down automaton with a finite stack size. In previous applications, maximum-likelihood estimation (MLE) is used to derive the parameters of the HVS model. However, MLE makes a number of assumptions and unfortunately some of these assumptions do not hold. Discriminative training, without making such assumptions, can improve the performance of the HVS model by discriminating the correct hypothesis from the competing hypotheses. Experiments have been conducted in two domains: the travel domain for the semantic parsing task using the DARPA Communicator data and the Air Travel Information Services (ATIS) data and the bioinformatics domain for the information extraction task using the GENIA corpus. The results demonstrate modest improvements of the performance of the HVS model using discriminative training. In the travel domain, discriminative training of the HVS model gives a relative error reduction rate of 31 percent in F-measure when compared with MLE on the DARPA Communicator data and 9 percent on the ATIS data. In the bioinformatics domain, a relative error reduction rate of 4 percent in F-measure is achieved on the GENIA corpus.

Learning sentiment classification model from labeled features

We propose a novel framework where an initial classifier is learned by incorporating prior information extracted from an existing sentiment lexicon. Preferences on expectations of sentiment labels of those lexicon words are expressed using generalized expectation criteria. Documents classified with high confidence are then used as pseudo-labeled examples for automatical domain-specific feature acquisition. The word-class distributions of such self-learned features are estimated from the pseudo-labeled examples and are used to train another classifier by constraining the model's predictions on unlabeled instances. Experiments on both the movie review data and the multi-domain sentiment dataset show that our approach attains comparable or better performance than exiting weakly-supervised sentiment classification methods despite using no labeled documents.

Differential effects of value consciousness and coupon proneness on consumers’ persuasion knowledge of pricing tactics

Pricing tactics persuasion knowledge (PTPK) is a relatively new concept that seeks to extend the research on persuasion knowledge to the pricing domain. Pricing tactics persuasion knowledge refers to the persuasion knowledge of consumers about marketers’ pricing tactics. Employing an acquisition–transaction utility theoretic perspective, this study examines the differential effects of value consciousness and coupon proneness on the accuracy, confidence, and calibration of consumers’ pricing tactics persuasion knowledge. The study finds that coupon proneness is negatively related to accuracy, confidence, and calibration of PTPK, while value consciousness is positively related to accuracy, confidence, and calibration of PTPK. The implications of the study are outlined.

Joint sentiment/topic model for sentiment analysis

Sentiment analysis or opinion mining aims to use automated tools to detect subjective information such as opinions, attitudes, and feelings expressed in text. This paper proposes a novel probabilistic modeling framework based on Latent Dirichlet Allocation (LDA), called joint sentiment/topic model (JST), which detects sentiment and topic simultaneously from text. Unlike other machine learning approaches to sentiment classification which often require labeled corpora for classifier training, the proposed JST model is fully unsupervised. The model has been evaluated on the movie review dataset to classify the review sentiment polarity and minimum prior information have also been explored to further improve the sentiment classification accuracy. Preliminary experiments have shown promising results achieved by JST.

Physicians' perceptions, attitudes and expectations regarding the role of hospital-based pharmacists in the West Bank, Palestine

Objective - To evaluate the perceptions, expectations and experiences of physicians with regard to hospital-based pharmacists in the West Bank, Palestine. Methods - A self-administered questionnaire was distributed to 250 physicians practising in four general hospitals in the West Bank, Palestine. The main sections of the questionnaire comprised a series of statements pertaining to physicians' perceptions, expectations and experiences with pharmacists. Key findings - One hundred and fifty seven questionnaires were completed and returned (response rate, 62.8%). The majority of respondents were most comfortable with pharmacists detecting and preventing prescription errors (76.4%; 95% confidence interval (CI) 69.5–81.2%) and patient education (57.9%; CI 51.2–63.4%) but they were not comfortable with pharmacists suggesting the use of prescription medications to patients (56.7%; CI 49.8–62.4%). Most physicians (62.4%; CI 56.8–69.1%) expected the pharmacist to educate their patients about the safe and appropriate use of their medication. However, approximately one-third (31.7%; CI 26.0–39.6%) did not expect pharmacists to be available for consultation during rounds. Physicians' experiences with pharmacists were less favourable; whereas 77% (CI 70.2–81.5%) of the physicians agreed that pharmacists were always a reliable source of information, only 11.5% (CI 6.2–16.4%) agreed that pharmacists appeared to be willing to take responsibility for solving any drug-related problems. Conclusion -The present study showed that hospital physicians are more likely to accept traditional pharmacy services than newer clinical services for hospital-based pharmacists in the West Bank, Palestine. Pharmacists should therefore interact more positively and more frequently with physicians. This will close the gap between the physicians' commonly held perceptions of what they expect pharmacists to do and what pharmacists can actually do, and gain support for an extended role of hospital-based pharmacists in future patient therapy management.

Subjective quality evaluation for voice communication over MANET

Due to the dynamic and mutihop nature of the Mobile Ad-hoc Network (MANET), voice communication over MANET may encounter many challenges. We set up a subjective quality evaluation model using ITU-T E-model with extension. And through simulation in NS-2, we evaluate how the following factors impact voice quality in MANET: the number of hops, the number of route breakages, the number of communication pairs and the background traffic. Using AODV as the underlying routing protocol, and with the MAC layer changed from 802.11 DCF to 802.11e EDCF, we observe that 802.11e is more suitable for implementating voice communication over MANET. © 2005 IEEE.

The attitudes of students and academic staff towards electronic course support:are we convergent?

The present study investigates the views and attitudes of both the students and staff with regard to the usefulness of electronic course support throughout all four years of the MPharm programme at Aston University. Students were sampled between January and March 2001 using a self-completion questionnaire administered during the start of a practical or tutorial class. All internal academic staff were interviewed using a semi-structured interview format. Response rates were 100 and 89.5%, respectively. The study found that students rapidly embraced the use of electronic course support within the undergraduate programme, although they view its role as augmenting traditional course delivery. This view was mirrored by the academic staff, although only around a half currently place their material on the University's virtual learning environment (VLE), WebCT. The failure of staff to completely embrace the VLE is grounded in a lack of confidence and ability in its use. A majority of the academic staff indicated that they wish to be trained further in the use of information technology. Academic institutions need to understand and meet these needs in parallel with the introduction of any electronic course support.

Exploring attitudes towards dietary modification and nutritional supplementation in people with and without age-related macular degeneration

Age-related macular degeneration (AMD) is the leading cause for visual impairment and blindness registration in the developed world. Due to the large amounts of conflicting AMD research on the role of nutrition and antioxidant intake, it is difficult for patients and practitioners to determine which measures can be taken to slow down the disease progression. The aim of this research was to determine the beliefs and knowledge that patients with AMD have about nutrition, to identify whether their condition is preventing them from eating a healthy diet, and to discover what their diet consists of. For the initial study, 158 participants with AMD (mean age 79 ± 7.8 years) and 50 participants without AMD (mean age 67 ± 8 years) were recruited from the Macular Society helpline, or from optometric practice. Participants had a 25 minute telephone interview where a 36-question survey was completed. The survey elicited demographic information, and questions covered the knowledge that participants had on nutrition and their current diet. The results from this survey uncovered three major findings: 1) 100% of AMD participants felt that they do not have enough information and support from eye-care practitioners regarding nutrition, 2) AMD patients are confused over, and display a lack of knowledge of, which foods are beneficial for eye health and when and what nutritional supplements to take, evidenced by 65% of participants not taking the correct dosage 3) AMD patients are not eating enough nutrients that would be beneficial for their condition - consuming an average of 1.4mg of lutein and zeaxanthin rather than the recommended 10mg. A clinical decision-making aid was created as an intervention based upon these findings. The aim of the aid was to help eye-care practitioners give the correct nutritional advice to their patients. Founded on the AREDS 2 inclusion and exclusion criteria, practitioners are able to identify which patients could benefit from a nutritional supplement, and which patients could benefit from dietary modification. An evaluation of the aid with 72 qualified eye-care practitioners exhibited a statistically significant increase in confidence after using the aid for two weeks. An evaluation using 51 student optometrists showed a statistically significant increase in confidence and a statistically significant increase in appropriate management of patients after using the aid. This project has elicited findings that are significant for AMD patient education. It is hoped that through these studies, patients will receive consistent advice about the risk factors for AMD, the link between AMD and nutrition, and the importance of maintaining a healthy, well-balanced diet.

Laser-induced breakdown spectroscopy quantitative analysis method via adaptive analytical line selection and relevance vector machine regression model

Abstract A new LIBS quantitative analysis method based on analytical line adaptive selection and Relevance Vector Machine (RVM) regression model is proposed. First, a scheme of adaptively selecting analytical line is put forward in order to overcome the drawback of high dependency on a priori knowledge. The candidate analytical lines are automatically selected based on the built-in characteristics of spectral lines, such as spectral intensity, wavelength and width at half height. The analytical lines which will be used as input variables of regression model are determined adaptively according to the samples for both training and testing. Second, an LIBS quantitative analysis method based on RVM is presented. The intensities of analytical lines and the elemental concentrations of certified standard samples are used to train the RVM regression model. The predicted elemental concentration analysis results will be given with a form of confidence interval of probabilistic distribution, which is helpful for evaluating the uncertainness contained in the measured spectra. Chromium concentration analysis experiments of 23 certified standard high-alloy steel samples have been carried out. The multiple correlation coefficient of the prediction was up to 98.85%, and the average relative error of the prediction was 4.01%. The experiment results showed that the proposed LIBS quantitative analysis method achieved better prediction accuracy and better modeling robustness compared with the methods based on partial least squares regression, artificial neural network and standard support vector machine.

Probabilistic multiple model neural network based leak detection system:experimental study

This paper presents an effective decision making system for leak detection based on multiple generalized linear models and clustering techniques. The training data for the proposed decision system is obtained by setting up an experimental pipeline fully operational distribution system. The system is also equipped with data logging for three variables; namely, inlet pressure, outlet pressure, and outlet flow. The experimental setup is designed such that multi-operational conditions of the distribution system, including multi pressure and multi flow can be obtained. We then statistically tested and showed that pressure and flow variables can be used as signature of leak under the designed multi-operational conditions. It is then shown that the detection of leakages based on the training and testing of the proposed multi model decision system with pre data clustering, under multi operational conditions produces better recognition rates in comparison to the training based on the single model approach. This decision system is then equipped with the estimation of confidence limits and a method is proposed for using these confidence limits for obtaining more robust leakage recognition results.

Improved Hybrid Model of HMM/GMM for Speech Recognition

In this paper, we propose a speech recognition engine using hybrid model of Hidden Markov Model (HMM) and Gaussian Mixture Model (GMM). Both the models have been trained independently and the respective likelihood values have been considered jointly and input to a decision logic which provides net likelihood as the output. This hybrid model has been compared with the HMM model. Training and testing has been done by using a database of 20 Hindi words spoken by 80 different speakers. Recognition rates achieved by normal HMM are 83.5% and it gets increased to 85% by using the hybrid approach of HMM and GMM.

Conditional Confidence Interval for the Scale Parameter of a Weibull Distribution

2000 Mathematics Subject Classification: 62F25, 62F03.

Em Estimation of the Offspring Distribution in Mutitype Branching Processes - a Model in Cell Kinetics

2000 Mathematics Subject Classification: 60J80, 60J85, 62P10, 92D25.

A Note on Bayesian Estimation for the Negative-Binomial Model

2000 Mathematics Subject Classification: 62F15.

Study on Robustness of Correlated Frailty Model

2000 Mathematics Subject Classifi cation: 62J12.