Author: LUIS ENRIQUE SUCAR SUCCAR
Artificial Intelligence applications in large-scale industry, such as fossil power plants, require the ability to manage uncertainty and time.
In this paper, we present an intelligent system to assist an operator of a power plant. This system, called SEDRET, is based on a novel
knowledge representation of uncertainty and time, called Temporal Nodes Bayesian Networks (TNBN), a type of Probabilistic Temporal
Network. A set of temporal nodes and a set of edge define a TNBN, each temporal node is defined by a value of a variable and a time interval
associate to the change of variable value. A TNBN generates a formal and systematic structure for modeling the temporal evolution of a
process under uncertainty. The inference mechanism is based on probabilistic reasoning. A TNBN can be used to recognize events and state
variables with respect to current plant conditions and predict the future propagation of disturbances. SEDRET was validated with the
diagnosis and prediction of events in a steam generator with a power plant training simulator. The results performed in this work indicate that
SEDRET can potentially improve plant availability through early diagnosis and prediction of disturbances that could lead to plant shutdown.
Temporal Nodes Bayesian Networks (TNBNs) and Networks of Probabilistic Events in Discrete Time (NPEDTs) are two different types of Event Bayesian Networks (EBNs). Both are based on the representation of uncertain events, alternatively to Dynamic Bayesian Networks, which deal with real-world dynamic properties. In a previous work, Arroyo-Figueroa and Sucar applied TNBNs to the diagnosis and prediction of the temporal faults that may occur in the steam generator of a fossil power plant. We present an NPEDT for the same domain, along with a comparative evaluation of the two networks. We examine different methods suggested in the literature for the evaluation of Bayesian networks, analyze their limitations when applied to this temporal domain, and suggest a new evaluation method appropriate for EBNs. In general, the results show that, in this domain, NPEDTs perform better than TNBNs, possibly due to be the finer time granularity used in the NPEDT.
Diagnosis and prediction m some
domains, like medical and industrial
diagnosis, require a representation that
combines uncertainty management and
temporal reasoning. Based on the fact
that in many cases there are few state
changes in the temporal range of
interest, we propose a novel representation
called Temporal Nodes Bayesian
Network (TNBN). In a TNBN each node
represents an event or state change of a
variable, and an arc corresponds to a
causal-temporal relation. The temporal
intervals can differ in number and size
for each temporal node, so this allows
multiple granularity. Our approach is
contrasted with a dynamic Bayesian
network for a simple medical example.
An empirical evaluation is presented for
a more complex problem, a subsystem of
a fossil power plant, in which this
approach is used for fault diagnosis and
event prediction with good results.
Dagnosis and prediction in some domains, like medicine, require and adequate representation taht combines uncertainty management and temporal reasoning. Novel representation called Temporal nodes Bayasean Network in which each node represents an event or state change of a variable, and an are corresponds to a causal-temporal relationship.
In arti®cial intelligence applications in large-scale industry, such as fossil fuel power plants, the knowledge about the process
comes from an expert's experience, and is generally expressed in a vague and fuzzy way, using ill-de®ned linguistic terms. This
paper presents a fuzzy intelligent system to assist an operator of fossil power plants. The approach is characterized as a fuzzy
diagnostic and fuzzy control system. The fuzzy diagnostic system is based on a novel representation for dealing with uncertainty
and time, called as fuzzy temporal network (FTN). An FTN is a formal and systematic structure, used to model temporal
linguistic sentences about the occurrence of an event. The fuzzy controller was designed for the regulation of the steam
temperature of a steam generator. The fuzzy rules were obtained by observing the dynamic characteristics of the steam
temperature response. The results show that the fuzzy controller has a better performance than advanced model-based
controller, either an dynamic matrix control (DMC) or a conventional PID controller. The main bene®ts are the reduction of
the overshoot and the tighter regulation of the superheater and reheater steam temperatures. The intelligent system has shown
that fuzzy logic techniques can play an important role in power-plant operation and control tasks. The scheme not only makes
the problem formulation more ¯exible but, if applied correctly, can improve the computational e ciency. This makes it practical
for many applications in complex ®elds where the real-time tasks are important.
In several domains it is common to have data from different, but closely related problems. For instance, in manufacturing, many products follow the same industrial process but with different conditions; or in industrial diagnosis, where there is equipment with similar specifications. In these cases it is common to have plenty of data for some scenarios but very little for others. In order to learn accurate models for rare cases, it is desirable to use data and knowledge from similar cases; a technique known as transfer learning. In this paper we propose an inductive transfer learning method for Bayesian networks, that considers both structure and parameter learning. For structure learning we use conditional independence tests, by combining measures from the target task with those obtained from one or more auxiliary tasks, using a novel weighted sum of the conditional independence measures. For parameter learning, we propose two variants of the linear pool for probability aggregation, combining the probability estimates from the target task with those from the auxiliary tasks. To validate our approach, we used three Bayesian networks models that are commonly used for evaluating learning techniques, and generated variants of each model by changing the structure as well as the parameters. We then learned one of the variants with a small dataset and combined it with information from the other variants. The experimental results show a significant improvement in terms of structure and parameters when we transfer knowledge from similar tasks. We also evaluated the method with real-world data from a manufacturing process considering several products, obtaining an improvement in terms of log-likelihood between the data and the model when we do transfer learning from related products.
The introduction of the notion of family resemblance represented a major shift in Wittgenstein’s thoughts on the meaning of words, moving away from a belief that words were well defined, to a view that words denoted less well defined categories of meaning. This paper presents the use of the notion of family resemblance in the area of machine learning as an example of the benefits that can accrue from adopting the kind of paradigm shift taken by Wittgenstein. The paper presents a model capable of learning exemplars using the principle of family resemblance and adopting Bayesian networks for a representation of exemplars. An empirical evaluation is presented on three data sets and shows promising results that suggest that previous assumptions about the way we categories need reopening.
Artificial Intelligence applications in large-scale industry, such as fossil fuel power plants, require the ability to manage
uncertainty and time. In these domains, the knowledge about the process comes from experts' experience and it is generally
expressed in a vague-fuzzy way using ill-defined linguistic terms. In this paper, we present a fuzzy expert system shell to assist an
operator of fossil power plants. The fuzzy expert system shell, called SADEP, is based on a new methodology for dealing with
uncertainty and time called Fuzzy Temporal Network (FTN). The FTN generates a formal and systematic structure used to model
the temporal evolution of a process under uncertainty. The inference mechanism for a FTN consists in the calculation of the
possibility degree of the real time occurrence of the events using the fuzzy compositional rule Sup_min, A FTN can be used to
recognize the significance of events and state variables with respect to current plant conditions and predict the future propagation
of disturbances. SADEP was validated with the diagnosis of two detailed disturbances of a fossil power plant: a power load
increment in the drum level and a water condenser pump failure. The evaluations performed in this work indicate that SADEP can
potentially improve plant availability through early diagnosis of disturbances that could lead to plant shutdown.
We have developed an affective behavior model for intelligent tutoring systems that considers both the affective and knowledge state of the student to generate tutorial actions. The affective behavior model was designed based on teachers expertise obtained through a survey which 11 math teachers participated. The study focused in knowing how teachers manage the affective state of the students in order the students learn. During the survey, teachers watched a video of a student interacting with an educational game with an animated pedagogical agent. We asked them which agent s animation and which pedagogical actions are suitable for affect and knowledge of the student in each student s movement.
This paper introduces two novel strategies for representing multimodal images with application to multimedia image retrieval. We consider images that are composed of both text and labels: while text describes the image content at a very high semantic level (e.g., making reference to places, dates or events), labels provide a mid-level description of the image (i.e., in terms of the objects that can be seen in the image). Accordingly, the main assumption of this work is that by combining information from text and labels we can develop very effective retrieval methods. We study standard information fusion techniques for combining both sources of information. However, whereas the performance of such techniques is highly competitive, they cannot capture effectively the content of images. Therefore, we propose two novel representations for multimodal images that attempt to exploit the semantic cohesion among terms from different modalities. Such representations are based on distributional term representations widely used in computational linguistics. Under the considered representations the content of an image is modeled by a distribution of co-occurrences over terms or of occurrences over other images, in such a way that the representation can be considered an expansion of the multimodal terms in the image. We report experimental results using the SAIAPR TC12 benchmark on two sets of topics used in ImageCLEF competitions with manually and automatically generated labels. Experimental results show that the proposed representations outperform significantly both, standard multimodal techniques and unimodal methods. Results on manually assigned labels provide an upper bound in the retrieval performance that can be obtained, whereas results with automatically generated labels are encouraging. The novel representations are able to capture more effectively the content of multimodal images. We emphasize that although we have applied our representations to multimedia image retrieval the same formulation can be adopted for modeling other multimodal documents (e.g., videos).
Multimedia image retrieval Image annotation Distributional term representations Semantic cohesion modeling CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA MATEMÁTICAS CIENCIA DE LOS ORDENADORES CIENCIA DE LOS ORDENADORES