Ramón García-Martínez

An Integrated Approach of Learning, Planning, and Execution

Agents (hardware or software) that act autonomously in an environment have to be able to integrate three basic behaviors: planning, execution, and learning. This integration is mandatory when the agent has no knowledge about how its actions can affect the environment, how the environment reacts to its actions, or, when the agent does not receive as an explicit input, the goals it must achieve. Without an “a priori” theory, autonomous agents should be able to self-propose goals, set-up plans for achieving the goals according to previously learned models of the agent and the environment, and learn those models from past experiences of successful and failed executions of plans. Planning involves selecting a goal to reach and computing a set of actions that will allow the autonomous agent to achieve the goal. Execution deals with the interaction with the environment by application of planned actions, observation of resulting perceptions, and control of successful achievement of the goals. Learning is needed to predict the reactions of the environment to the agent actions, thus guiding the agent to achieve its goals more efficiently.In this context, most of the learning systems applied to problem solving have been used to learn control knowledge for guiding the search for a plan, but few systems have focused on the acquisition of planning operator descriptions. As an example, currently, one of the most used techniques for the integration of (a way of) planning, execution, and learning is reinforcement learning. However, they usually do not consider the representation of action descriptions, so they cannot reason in terms of goals and ways of achieving those goals.In this paper, we present an integrated architecture, lope, that learns operator definitions, plans using those operators, and executes the plans for modifying the acquired operators. The resulting system is domain-independent, and we have performed experiments in a robotic framework. The results clearly show that the integrated planning, learning, and executing system outperforms the basic planner in that domain.

Information mining processes based on intelligent systems

Business Intelligence offers an interdisciplinary approach (within which is Information Systems), that taking all available information resources and using of analytical and synthesis tools with the ability to transform information into knowledge, focuses on generating knowledge that contributes to the management decision-making and generation of strategic plans in organizations. Information Mining is the sub-discipline of information systems which supports business intelligence tools to transform information into knowledge. It has defined as the search for interesting patterns and important regularities in large bodies of information. We address the need to identify information mining processes to obtain knowledge from available information. When information mining processes are defined, we may decide which data mining algorithms will support the information mining processes. In this context, this paper proposes a characterization of the information mining process related to the following business intelligence problems: discovery of rules of behavior, discovery of groups, discovery of significant attributes, discovering rules of group membership and weight of rules of behavior or rules of group memberships.

Fuzzy Control For Improving Energy Management Within Indoor Building Environments

Research on energy efficiency in buildings has proved that the design and even the sophisticated facilities such as building energy management systems (BEMS), aiming to improve the indoor environmental conditions while minimizing the energy needs, are not sufficient enough due to users interference. Users are a dynamic part of the building, therefore they should be taken into account in the control strategy. Latest trends in designing intelligent building energy management systems (IBEMS) integrate a man machine interface that could store the users preferences and adapt the control strategy accordingly. The objective of the present paper is to present the advantages of the fuzzy control techniques together with a man machine interface based on a smart card terminal in satisfying the user preferences. A fuzzy controller is developed and the minimization of energy consumption is achieved by the use of a suitable cost function for the whole system. The overall control system including the cost function has been modeled and tested by using the software tool MATLAB/SIMULINK.

PREDICTION IN HEALTH DOMAIN USING BAYESIAN NETWORKS OPTIMIZATION BASED ON INDUCTION LEARNING TECHNIQUES

A Bayesian network is a directed acyclic graph in which each node represents a variable and each arc a probabilistic dependency; they are used to provide: a compact form to represent the knowledge and flexible methods of reasoning. Obtaining it from data is a learning process that is divided in two steps: structural learning and parametric learning. In this paper we define an automatic learning method that optimizes the Bayesian networks applied to classification, using a hybrid method of learning that combines the advantages of the induction techniques of the decision trees (TDIDT-C4.5) with those of the Bayesian networks. The resulting method is applied to prediction in health domain.

Patterns in temporal series of meteorological variables using SOM & TDIDT

The purpose of the present article is to investigate if there exist any such set of temporal stable patterns in temporal series of meteorological variables studying series of air temperature, wind speed and direction an atmospheric pressure in a period with meteorological conditions involving nocturnal inversion of air temperature in Allen, Rio Negro, Argentina. Our conjecture is that there exist independent stable temporal activities, the mixture of which give rise to the weather variables; and these stable activities could be extracted by Self Organized Maps plus Top Down Induction Decision Trees analysis of the data arising from the weather patterns, viewing them as temporal signals.

Outlier detection in audit logs for application systems

Abstract An outlier is defined as an observation that is significantly different from the other data in its set. An auditor will employ many techniques, processes and tools to identify these entries, and data mining is one such medium through which the auditor can analyze information. The enormous amount of information contained within transactional processing systems׳ logs means that auditors must employ automated systems for anomalous data detection. Several data mining algorithms have been tested, especially those that deal specifically with classification and outlier detection. A group of these previously described algorithms was selected for use in designing and developing a process to assist the auditor in anomalous data detection within audit logs. We have been successful in creating and ratifying an outlier detection process that works in the alphanumeric fields of the audit logs from an information system, thus constituting a useful tool for system auditors performing data analysis tasks.

A Proposal of Interaction Modelling Formalisms in Virtual Collaborative Work Spaces

The virtual collaborative workspaces allow integration of work groups in which members are not physically contiguous. There is a vast literature related to modeling of software architectures that support this type of environment. However, the existing formalisms model the interaction between actors and between system and system components. In this context, this paper proposes modeling artifacts (which supplement the existing) to describe: domain vocabulary through table category-concept-definition; interactions among actors through interaction cases, interaction group diagrams, interaction procedures and sequence diagram of group dynamics; and development by group of conceptual artifacts through diagram of development of conceptual objects. we present a proof of concept to illustrate the proposal.

Deriving Processes of Information Mining Based on Semantic Nets and Frames

There are information mining methodologies that emphasize the importance of planning for requirements elicitation along the entire project in an orderly, documented, consistent and traceable manner. However, given the characteristics of this type of project, the approach proposed by the classical requirements engineering is not applicable to the process of identifying the problem of information mining, nor allows to infer from the business domain modelling, the information mining process which solves it. This paper proposes an extension of semantic nets and frames to represent knowledge of the business domain, business problem and problem of information mining; and a methodology to derive the information mining process from the proposed knowledge representations is introduced.

Collaborative Research Training Based on Virtual Spaces

A possible strategy for training researchers is to provide integrated research cores with researchers-in-training under the guidance of a senior trained researcher. Information technology and communication have enabled the construction of virtual communities formed by individuals who may be far away physically but who are cognitively close, hence giving rise to collaborative research training models. In this context, this paper formulates an approach to identify the elements of the work space of a research group devoted to research training and to assess the technological feasibility of virtualization of such elements.

Feasibility and effort estimation models for medium and small size information mining projects

Abstract Information mining is a sub-discipline of Information Systems which provides the non-trivial knowledge needed for making decisions inside an organization. Although such projects have different features from Software Engineering ones, they share some of their problems. Among these problems two are highlighted: unmanaged risks and inaccurate estimations of necessary resources to complete the project. In this context, this paper presents two ad-hoc models to be applied in Small and Medium-sized Enterprises: one for assessing project feasibility and the other for estimating the resources and time required to carry out the project. Both models should be applied at the beginning of the project.