Search results for: causal inference

Authors: Tharini N. de Silva, Xiao Zhibo, Zhao Rui, Mao Kezhi

Abstract:

Causal relation identification is a crucial task in information extraction and knowledge discovery. In this work, we present two approaches to causal relation identification. The first is a classification model trained on a set of knowledge-based features. The second is a deep learning based approach training a model using convolutional neural networks to classify causal relations. We experiment with several different convolutional neural networks (CNN) models based on previous work on relation extraction as well as our own research. Our models are able to identify both explicit and implicit causal relations as well as the direction of the causal relation. The results of our experiments show a higher accuracy than previously achieved for causal relation identification tasks.

Keywords: Causal relation identification, convolutional neural networks, natural Language Processing, Machine Learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2257

Authors: Federico Cabitza, Bernardo Dal Seno

Abstract:

In this paper DJess is presented, a novel distributed production system that provides an infrastructure for factual and procedural knowledge sharing. DJess is a Java package that provides programmers with a lightweight middleware by which inference systems implemented in Jess and running on different nodes of a network can communicate. Communication and coordination among inference systems (agents) is achieved through the ability of each agent to transparently and asynchronously reason on inferred knowledge (facts) that might be collected and asserted by other agents on the basis of inference code (rules) that might be either local or transmitted by any node to any other node.

Keywords: Knowledge-Based Systems, Expert Systems, Distributed Inference Systems, Parallel Production Systems, Ambient Intelligence, Mobile Agents.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1796

Authors: May George

Abstract:

Inference plays an important role in the learning process and it can lead to a rapid acquisition of a second language. When learning a non-native language i.e., a critical language like Arabic, the students depend on the teacher’s support most of the time to learn new concepts. The students focus on memorizing the new vocabulary and stress on learning all the grammatical rules. Hence, the students became mechanical and cannot produce the language easily. As a result, they are unable to predicate the meaning of words in the context by relying heavily on the teacher, in that they cannot link their prior knowledge or even identify the meaning of the words without the support of the teacher. This study explores how the teacher guides students learning during the inference process and what are the processes of learning that can direct student’s inference.

Keywords: Inference, Reading, Arabic, and Language Acquisition.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2052

Authors: Alicia Y. C. Tang, Rukaini Abdullah, Sharifuddin M. Zain, Noorsaadah A. Rahman

Abstract:

This paper discusses the causal explanation capability of QRIOM, a tool aimed at supporting learning of organic chemistry reactions. The development of the tool is based on the hybrid use of Qualitative Reasoning (QR) technique and Qualitative Process Theory (QPT) ontology. Our simulation combines symbolic, qualitative description of relations with quantity analysis to generate causal graphs. The pedagogy embedded in the simulator is to both simulate and explain organic reactions. Qualitative reasoning through a causal chain will be presented to explain the overall changes made on the substrate; from initial substrate until the production of final outputs. Several uses of the QPT modeling constructs in supporting behavioral and causal explanation during run-time will also be demonstrated. Explaining organic reactions through causal graph trace can help improve the reasoning ability of learners in that their conceptual understanding of the subject is nurtured.

Keywords: Qualitative reasoning, causal graph, organicreactions, explanation, QPT, modeling constructs.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1419

Authors: Luis A. Morales Rosales, Saul E. Pomares Hernandez, Gustavo Rodriguez Gomez

Abstract:

Synchronous cooperative systems (SCS) bring together users that are geographically distributed and connected through a network to carry out a task. Examples of SCS include Tele- Immersion and Tele-Conferences. In SCS, the coordination is the core of the system, and it has been defined as the act of managing interdependencies between activities performed to achieve a goal. Some of the main problems that SCS present deal with the management of constraints between simultaneous activities and the execution ordering of these activities. In order to resolve these problems, orderings based on Lamport-s happened-before relation have been used, namely, causal, Δ-causal, and causal-total orderings. They mainly differ in the degree of asynchronous execution allowed. One of the most important orderings is the causal order, which establishes that the events must be seen in the cause-effect order as they occur in the system. In this paper we show that for certain SCS (e.g. videoconferences, tele-immersion) where some degradation of the system is allowed, ensuring the causal order is still rigid, which can render negative affects to the system. In this paper, we illustrate how a more relaxed ordering, which we call Fuzzy Causal Order (FCO), is useful for such kind of systems by allowing a more asynchronous execution than the causal order. The benefit of the FCO is illustrated by applying it to a particular scenario of intermedia synchronization of an audio-conference system.

Keywords: Event ordering, fuzzy causal ordering, happenedbefore relation and cooperative systems.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1497

Authors: J. Fernandez, N. Aguilar, R. Fernandez de Canete, J. C. Ramos-Diaz

Abstract:

In this paper, a simulation model of the glucose-insulin system for a patient undergoing diabetes Type 1 is developed by using a causal modeling approach under system dynamics. The OpenModelica simulation environment has been employed to build the so called causal model, while the glucose-insulin model parameters were adjusted to fit recorded mean data of a diabetic patient database. Model results under different conditions of a three-meal glucose and exogenous insulin ingestion patterns have been obtained. This simulation model can be useful to evaluate glucose-insulin performance in several circumstances, including insulin infusion algorithms in open-loop and decision support systems in closed-loop.

Keywords: Causal modeling, diabetes, glucose-insulin system, diabetes, causal modeling, OpenModelica software.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1425

Authors: Josephine S. Larkings, Patricia M. Brown

Abstract:

Mental health professionals views about mental illness is an important issue which has not received enough attention. The negative stigma associated with mental illness can have many negative consequences. Unfortunately, health professionals working with the mentally ill can also exhibit stigma. It has been suggested that causal explanations or beliefs around the causes of mental illness may influence stigma. This study aims to gain a greater insight into stigma through examining stigma among potential mental health professionals. Firstly, results found that potential mental health professionals had relatively low social distance t(205) = -3.62, p <.001. Secondly, an ANOVA indicated that the participants endorsed some causal beliefs more than others, F(1.82, 311.55) = 88.47, p < .001, partial n2 = .34. Moreover, participants endorsed the biological causal explanation the most. Thirdly, results indicated that combined contact (quality and quantity) and causal beliefs (biological, psychological, and environmental) explained a significant proportion of the variance in stigma, R2 = .35, adjusted R2 = .33, F(5, 153) = 16.66, p < .001. Quality of contact was the strongest predictor, with greater quality of contact associated with lower desired social distance. Also, quantity of contact, psychological and environmental causal explanations were also significant predictors of stigma. Greater quantity of contact and higher levels of environmental causal beliefs were associated with lower levels of stigma while psychological causal explanations were associated with higher levels of stigma. A series of multiple regression analyses were conducted that showed the three causal beliefs had different impacts on four factors of stigma (Authoritarian, Benevolence, Social restrictiveness, and Community mental health ideology). These analyses showed that psychological causal beliefs had the most positive impact. More research is required on this topic as it has important implications to the treatment and recovery for people suffering from mental illness.

Keywords: Causal beliefs, contact, health professionals, mental illness stigma, social distance.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2074

Authors: Azhar Rauf, Sareer Badshah, Shah Khusro

Abstract:

This paper analyzes different techniques of the fine grained security of relational databases for the two variables-data accessibility and inference. Data accessibility measures the amount of data available to the users after applying a security technique on a table. Inference is the proportion of information leakage after suppressing a cell containing secret data. A row containing a secret cell which is suppressed can become a security threat if an intruder generates useful information from the related visible information of the same row. This paper measures data accessibility and inference associated with row, cell, and column level security techniques. Cell level security offers greatest data accessibility as it suppresses secret data only. But on the other hand, there is a high probability of inference in cell level security. Row and column level security techniques have least data accessibility and inference. This paper introduces cell plus innocent security technique that utilizes the cell level security method but suppresses some innocent data to dodge an intruder that a suppressed cell may not necessarily contain secret data. Four variations of the technique namely cell plus innocent 1/4, cell plus innocent 2/4, cell plus innocent 3/4, and cell plus innocent 4/4 respectively have been introduced to suppress innocent data equal to 1/4, 2/4, 3/4, and 4/4 percent of the true secret data inside the database. Results show that the new technique offers better control over data accessibility and inference as compared to the state-of-theart security techniques. This paper further discusses the combination of techniques together to be used. The paper shows that cell plus innocent 1/4, 2/4, and 3/4 techniques can be used as a replacement for the cell level security.

Keywords: Fine Grained Security, Data Accessibility, Inference, Row, Cell, Column Level Security.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1471

Authors: Yuanyuan Chai, Limin Jia, Zundong Zhang

Abstract:

Hybrid algorithm is the hot issue in Computational Intelligence (CI) study. From in-depth discussion on Simulation Mechanism Based (SMB) classification method and composite patterns, this paper presents the Mamdani model based Adaptive Neural Fuzzy Inference System (M-ANFIS) and weight updating formula in consideration with qualitative representation of inference consequent parts in fuzzy neural networks. M-ANFIS model adopts Mamdani fuzzy inference system which has advantages in consequent part. Experiment results of applying M-ANFIS to evaluate traffic Level of service show that M-ANFIS, as a new hybrid algorithm in computational intelligence, has great advantages in non-linear modeling, membership functions in consequent parts, scale of training data and amount of adjusted parameters.

Keywords: Fuzzy neural networks, Mamdani fuzzy inference, M-ANFIS

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5244

Authors: Abdelaziz Zaidi, Belkacem Ould Bouamama, Moncef Tagina

Abstract:

Bond Graph as a unified multidisciplinary tool is widely used not only for dynamic modelling but also for Fault Detection and Isolation because of its structural and causal proprieties. A binary Fault Signature Matrix is systematically generated but to make the final binary decision is not always feasible because of the problems revealed by such method. The purpose of this paper is introducing a methodology for the improvement of the classical binary method of decision-making, so that the unknown and identical failure signatures can be treated to improve the robustness. This approach consists of associating the evaluated residuals and the components reliability data to build a Hybrid Bayesian Network. This network is used in two distinct inference procedures: one for the continuous part and the other for the discrete part. The continuous nodes of the network are the prior probabilities of the components failures, which are used by the inference procedure on the discrete part to compute the posterior probabilities of the failures. The developed methodology is applied to a real steam generator pilot process.

Keywords: Redundancy relations, decision-making, Bond Graph, reliability, Bayesian Networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2525

Authors: A. Taqaddas

Abstract:

This mini literature review discusses casual relationships between HIV and AIDS in view of Koch’s postulates.

Keywords: AIDS, Causal Relationship, HIV, Koch’s Postulates.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2782

Authors: Adel Aloraini

Abstract:

Financial forecasting using machine learning techniques has received great efforts in the last decide . In this ongoing work, we show how machine learning of graphical models will be able to infer a visualized causal interactions between different banks in the Saudi equities market. One important discovery from such learned causal graphs is how companies influence each other and to what extend. In this work, a set of graphical models named Gaussian graphical models with developed ensemble penalized feature selection methods that combine ; filtering method, wrapper method and a regularizer will be shown. A comparison between these different developed ensemble combinations will also be shown. The best ensemble method will be used to infer the causal relationships between banks in Saudi equities market.

Keywords: Causal interactions , banks, feature selection, regularizere,

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1747

Authors: Shabbar Naqvi, Jonathan M. Garibaldi

Abstract:

Breast cancer is one of the most frequent occurring cancers in women throughout the world including U.K. The grading of this cancer plays a vital role in the prognosis of the disease. In this paper we present an overview of the use of advanced computational method of fuzzy inference system as a tool for the automation of breast cancer grading. A new spectral data set obtained from Fourier Transform Infrared Spectroscopy (FTIR) of cancer patients has been used for this study. The future work outlines the potential areas of fuzzy systems that can be used for the automation of breast cancer grading.

Keywords: Breast cancer, FTIR, fuzzy inference system, principal component analysis

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2130

Authors: Pavan S., Sridhar G., Sridhar V.

Abstract:

Super resolution is one of the commonly referred inference problems in computer vision. In the case of images, this problem is generally addressed using a graphical model framework wherein each node represents a portion of the image and the edges between the nodes represent the statistical dependencies. However, the large dimensionality of images along with the large number of possible states for a node makes the inference problem computationally intractable. In this paper, we propose a representation wherein each node can be represented as acombination of multiple regression functions. The proposed approach achieves a tradeoff between the computational complexity and inference accuracy by varying the number of regression functions for a node.

Keywords: Belief propagation, Graphical model, Regression, Super resolution.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1547

Authors: Yilun Shang

Abstract:

We consider a heterogeneously mixing SIR stochastic epidemic process in populations described by a general graph. Likelihood theory is developed to facilitate statistic inference for the parameters of the model under complete observation. We show that these estimators are asymptotically Gaussian unbiased estimates by using a martingale central limit theorem.

Keywords: statistic inference, maximum likelihood, epidemicmodel, heterogeneous mixing.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1409

Authors: Mazliham Mohd Su'ud, Pierre Loonis, Idris Abu Seman

Abstract:

This paper deals with the extraction of information from the experts to automatically identify and recognize Ganoderma infection in oil palm stem using tomography images. Expert-s knowledge are used as rules in a Fuzzy Inference Systems to classify each individual patterns observed in he tomography image. The classification is done by defining membership functions which assigned a set of three possible hypotheses : Ganoderma infection (G), non Ganoderma infection (N) or intact stem tissue (I) to every abnormalities pattern found in the tomography image. A complete comparison between Mamdani and Sugeno style,triangular, trapezoids and mixed triangular-trapezoids membership functions and different methods of aggregation and defuzzification is also presented and analyzed to select suitable Fuzzy Inference System methods to perform the above mentioned task. The results showed that seven out of 30 initial possible combination of available Fuzzy Inference methods in MATLAB Fuzzy Toolbox were observed giving result close to the experts estimation.

Keywords: Fuzzy Inference Systems, Tomography analysis, Modelizationof expert's information, Ganoderma Infection pattern recognition

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1836

Authors: J. Hossen, A. Rahman, K. Samsudin, F. Rokhani, S. Sayeed, R. Hasan

Abstract:

The Neuro-Fuzzy hybridization scheme has become of research interest in pattern classification over the past decade. The present paper proposes a novel Modified Adaptive Fuzzy Inference Engine (MAFIE) for pattern classification. A modified Apriori algorithm technique is utilized to reduce a minimal set of decision rules based on input output data sets. A TSK type fuzzy inference system is constructed by the automatic generation of membership functions and rules by the fuzzy c-means clustering and Apriori algorithm technique, respectively. The generated adaptive fuzzy inference engine is adjusted by the least-squares fit and a conjugate gradient descent algorithm towards better performance with a minimal set of rules. The proposed MAFIE is able to reduce the number of rules which increases exponentially when more input variables are involved. The performance of the proposed MAFIE is compared with other existing applications of pattern classification schemes using Fisher-s Iris and Wisconsin breast cancer data sets and shown to be very competitive.

Keywords: Apriori algorithm, Fuzzy C-means, MAFIE, TSK

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1931

Authors: Eleftherios Giovanis

Abstract:

In this paper discrete choice models, Logit and Probit are examined in order to predict the economic recession or expansion periods in USA. Additionally we propose an adaptive neuro-fuzzy inference system with triangular membership function. We examine the in-sample period 1947-2005 and we test the models in the out-of sample period 2006-2009. The forecasting results indicate that the Adaptive Neuro-fuzzy Inference System (ANFIS) model outperforms significant the Logit and Probit models in the out-of sample period. This indicates that neuro-fuzzy model provides a better and more reliable signal on whether or not a financial crisis will take place.

Keywords: ANFIS, discrete choice models, financial crisis, USeconomy

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1610

Authors: Erika Martinez Ramirez, Rene V. Mayorga

Abstract:

Nowadays, more engineering systems are using some kind of Artificial Intelligence (AI) for the development of their processes. Some well-known AI techniques include artificial neural nets, fuzzy inference systems, and neuro-fuzzy inference systems among others. Furthermore, many decision-making applications base their intelligent processes on Fuzzy Logic; due to the Fuzzy Inference Systems (FIS) capability to deal with problems that are based on user knowledge and experience. Also, knowing that users have a wide variety of distinctiveness, and generally, provide uncertain data, this information can be used and properly processed by a FIS. To properly consider uncertainty and inexact system input values, FIS normally use Membership Functions (MF) that represent a degree of user satisfaction on certain conditions and/or constraints. In order to define the parameters of the MFs, the knowledge from experts in the field is very important. This knowledge defines the MF shape to process the user inputs and through fuzzy reasoning and inference mechanisms, the FIS can provide an “appropriate" output. However an important issue immediately arises: How can it be assured that the obtained output is the optimum solution? How can it be guaranteed that each MF has an optimum shape? A viable solution to these questions is through the MFs parameter optimization. In this Paper a novel parameter optimization process is presented. The process for FIS parameter optimization consists of the five simple steps that can be easily realized off-line. Here the proposed process of FIS parameter optimization it is demonstrated by its implementation on an Intelligent Interface section dealing with the on-line customization / personalization of internet portals applied to E-commerce.

Keywords: Artificial Intelligence, Fuzzy Logic, Fuzzy InferenceSystems, Nonlinear Optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1984

Authors: Mou-Yuan Liao, Chien-Wei Wu, Chien-Hua Lin

Abstract:

Process capability index Cpk is the most widely used index in making managerial decisions since it provides bounds on the process yield for normally distributed processes. However, existent methods for assessing process performance which constructed by statistical inference may unfortunately lead to fine results, because uncertainties exist in most real-world applications. Thus, this study adopts fuzzy inference to deal with testing of Cpk . A brief score is obtained for assessing a supplier’s process instead of a severe evaluation.

Keywords: Process capability analysis, quality control.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1427

Authors: Sherin M. Youssef, Ahmed Abouelfarag, Noha M. Ghatwary

Abstract:

An adaptive Fuzzy Inference Perceptual model has been proposed for watermarking of digital images. The model depends on the human visual characteristics of image sub-regions in the frequency multi-resolution wavelet domain. In the proposed model, a multi-variable fuzzy based architecture has been designed to produce a perceptual membership degree for both candidate embedding sub-regions and strength watermark embedding factor. Different sizes of benchmark images with different sizes of watermarks have been applied on the model. Several experimental attacks have been applied such as JPEG compression, noises and rotation, to ensure the robustness of the scheme. In addition, the model has been compared with different watermarking schemes. The proposed model showed its robustness to attacks and at the same time achieved a high level of imperceptibility.

Keywords: Watermarking, The human visual system (HVS), Fuzzy Inference System (FIS), Local Binary Pattern (LBP), Discrete Wavelet Transform (DWT).

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1818

Authors: Ahmad Habibizad Navin, Mehdi Naghian Fesharaki, Mirkamal Mirnia, Mohamad Teshnelab, Ehsan Shahamatnia

Abstract:

In this paper we introduce new data oriented modeling of uniform random variable well-matched with computing systems. Due to this conformity with current computers structure, this modeling will be efficiently used in statistical inference.

Keywords: Uniform random variable, Data oriented modeling, Statistical inference, Prodigraph, Statistically complete tree, Uniformdigital probability digraph, Uniform n-complete probability tree.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1631

Authors: Ajay Shekhar Pandey, D. Singh, S. K. Sinha

Abstract:

A clustering based technique has been developed and implemented for Short Term Load Forecasting, in this article. Formulation has been done using Mean Absolute Percentage Error (MAPE) as an objective function. Data Matrix and cluster size are optimization variables. Model designed, uses two temperature variables. This is compared with six input Radial Basis Function Neural Network (RBFNN) and Fuzzy Inference Neural Network (FINN) for the data of the same system, for same time period. The fuzzy inference system has the network structure and the training procedure of a neural network which initially creates a rule base from existing historical load data. It is observed that the proposed clustering based model is giving better forecasting accuracy as compared to the other two methods. Test results also indicate that the RBFNN can forecast future loads with accuracy comparable to that of proposed method, where as the training time required in the case of FINN is much less.

Keywords: Load forecasting, clustering, fuzzy inference.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1626

Authors: Yohei Saika, Hiroki Sakaematsu, Shota Akiyama

Abstract:

We investigated statistical performance of Bayesian inference using maximum entropy and MAP estimation for several models which approximated wave-fronts in remote sensing using SAR interferometry. Using Monte Carlo simulation for a set of wave-fronts generated by assumed true prior, we found that the method of maximum entropy realized the optimal performance around the Bayes-optimal conditions by using model of the true prior and the likelihood representing optical measurement due to the interferometer. Also, we found that the MAP estimation regarded as a deterministic limit of maximum entropy almost achieved the same performance as the Bayes-optimal solution for the set of wave-fronts. Then, we clarified that the MAP estimation perfectly carried out phase unwrapping without using prior information, and also that the MAP estimation realized accurate phase unwrapping using conjugate gradient (CG) method, if we assumed the model of the true prior appropriately.

Keywords: Bayesian inference using maximum entropy, MAP estimation using conjugate gradient method, SAR interferometry.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1751

Authors: Andy Handouyahia, Tony Haddad, Frank Eaton

Abstract:

Recent quasi-experimental evaluation of the Canadian Active Labour Market Policies (ALMP) by Human Resources and Skills Development Canada (HRSDC) has provided an opportunity to examine alternative methods to estimating the incremental effects of Employment Benefits and Support Measures (EBSMs) on program participants. The focus of this paper is to assess the efficiency and robustness of inverse probability weighting (IPW) relative to kernel matching (KM) in the estimation of program effects. To accomplish this objective, the authors compare pairs of 1,080 estimates, along with their associated standard errors, to assess which type of estimate is generally more efficient and robust. In the interest of practicality, the authorsalso document the computationaltime it took to produce the IPW and KM estimates, respectively.

Keywords: Treatment effect, causal inference, observational studies, Propensity score based matching, Kernel Matching, Inverse Probability Weighting, Estimation methods for incremental effect.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 6925

Authors: Zhen Cheng, Xinyu Dai, Shujian Huang, Jiajun Chen

Abstract:

Recently, explanatory natural language inference has attracted much attention for the interpretability of logic relationship prediction, which is also known as explanation generation for Natural Language Inference (NLI). Existing explanation generators based on discriminative Encoder-Decoder architecture have achieved noticeable results. However, we find that these discriminative generators usually generate explanations with correct evidence but incorrect logic semantic. It is due to that logic information is implicitly encoded in the premise-hypothesis pairs and difficult to model. Actually, logic information identically exists between premise-hypothesis pair and explanation. And it is easy to extract logic information that is explicitly contained in the target explanation. Hence we assume that there exists a latent space of logic information while generating explanations. Specifically, we propose a generative model called Variational Explanation Generator (VariationalEG) with a latent variable to model this space. Training with the guide of explicit logic information in target explanations, latent variable in VariationalEG could capture the implicit logic information in premise-hypothesis pairs effectively. Additionally, to tackle the problem of posterior collapse while training VariaztionalEG, we propose a simple yet effective approach called Logic Supervision on the latent variable to force it to encode logic information. Experiments on explanation generation benchmark—explanation-Stanford Natural Language Inference (e-SNLI) demonstrate that the proposed VariationalEG achieves significant improvement compared to previous studies and yields a state-of-the-art result. Furthermore, we perform the analysis of generated explanations to demonstrate the effect of the latent variable.

Keywords: Natural Language Inference, explanation generation, variational auto-encoder, generative model.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 692

Authors: Ashok Kumar, Vinod Kumar

Abstract:

Wireless Sensor Networks (WSNs) are used to monitor/observe vast inaccessible regions through deployment of large number of sensor nodes in the sensing area. For majority of WSN applications, the collected data needs to be combined with geographic information of its origin to make it useful for the user; information received from remote Sensor Nodes (SNs) that are several hops away from base station/sink is meaningless without knowledge of its source. In addition to this, location information of SNs can also be used to propose/develop new network protocols for WSNs to improve their energy efficiency and lifetime. In this paper, range free localization protocols for WSNs have been proposed. The proposed protocols are based on weighted centroid localization technique, where the edge weights of SNs are decided by utilizing fuzzy logic inference for received signal strength and link quality between the nodes. The fuzzification is carried out using (i) Mamdani, (ii) Sugeno, and (iii) Combined Mamdani Sugeno fuzzy logic inference. Simulation results demonstrate that proposed protocols provide better accuracy in node localization compared to conventional centroid based localization protocols despite presence of unintentional radio frequency interference from radio frequency (RF) sources operating in same frequency band.

Keywords: localization, range free, received signal strength, link quality indicator, Mamdani fuzzy logic inference, Sugeno fuzzy logic inference.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2631

Authors: Rachel C. F. Sun

Abstract:

This study aimed to examine the similarities and differences between teachers’ and students’ causal explanations of classroom misbehavior. In-depth semi-structured interviews were conducted with twelve teachers and eighteen Grade 7-9 students. The qualitative data were analyzed, in which the attributed causes of classroom misbehavior were categorized into student, family, school and peer factors. Findings showed that both interviewed teachers and students shared similarity in attributing to student factors, such as ‘fun and pleasure seeking’ and ‘attention seeking’ as the leading causes of misbehavior. However, the students accounted to school factors, particularly ‘boring lessons’ as the next attributed causes, while the teachers accounted to family factors, such as ‘lack of parent demandingness’. By delineating the factors at student, family, school, and peer levels, these findings help drawing corresponding implications for preventing and mitigating misbehavior in school.

Keywords: Causal explanation, misbehavior, student, teacher.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5288

Authors: Amir Azizi, Amir Yazid B. Ali, Loh Wei Ping

Abstract:

Throughput is an important measure of performance of production system. Analyzing and modeling of production throughput is complex in today-s dynamic production systems due to uncertainties of production system. The main reasons are that uncertainties are materialized when the production line faces changes in setup time, machinery break down, lead time of manufacturing, and scraps. Besides, demand changes are fluctuating from time to time for each product type. These uncertainties affect the production performance. This paper proposes Bayesian inference for throughput modeling under five production uncertainties. Bayesian model utilized prior distributions related to previous information about the uncertainties where likelihood distributions are associated to the observed data. Gibbs sampling algorithm as the robust procedure of Monte Carlo Markov chain was employed for sampling unknown parameters and estimating the posterior mean of uncertainties. The Bayesian model was validated with respect to convergence and efficiency of its outputs. The results presented that the proposed Bayesian models were capable to predict the production throughput with accuracy of 98.3%.

Keywords: Bayesian inference, Uncertainty modeling, Monte Carlo Markov chain, Gibbs sampling, Production throughput

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2145

Authors: L. Richter, J. Lübkemann, P. Nyhuis

Abstract:

In order to remain competitive in what is a turbulent environment; businesses must be able to react rapidly to change. The past response to volatile market conditions was to introduce an element of flexibility to production. Nowadays, what is often required is a redesign of factory structures in order to cope with the state of constant flux. The Institute of Production Systems and Logistics is currently developing a descriptive and causal model for the redesign of plant structures as part of an ongoing research project. This article presents the first research findings attained in devising this model.

Keywords: Causal model, change driven factory redesign, factory planning, plant structure.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1815