Knowledge-Based Intelligent Information and Engineering Systems

Knowledge-Based and Intelligent Information and Engineering Systems

Knowledge-Based and Intelligent Information and Engineering Systems

Knowledge-Based Intelligent Information and Engineering Systems

There can be little doubt that training at altitude is fundamental to preparing an athlete for competition at altitude. However the value of training at altitude for competition at sea level appears on the one hand to lack total acceptance amongst sports scientists; and on the other to hold some cloak of mystery for coaches who have yet to enjoy first hand experience. The fact is that very few endurance athletes will ignore the critical edge which altitude training affords. Each fraction of a percentage of performance advantage gained through methods which are within the rules of fair play in sport, may shift the balance between failure and achievement. Moreover, there is growing support for application of training at altitude for speed-related disciplines. This paper aims to demystify the subject by dealing with practical aspects of training at altitude. Such aspects include a checklist of what should and should not be done at altitude, when to use altitude relative to target competitions, and specific training examples.

In social settings, people interact in close proximity. When analyzing such encounters from video, we are typically interested in distinguishing between a large number of different interactions. Here, we address training deformable part models (DPMs) for the detection of such interactions from video, in both space and time. When we consider a large number of interaction classes, we face two challenges. First, we need to distinguish between interactions that are visually more similar. Second, it becomes more difficult to obtain sufficient specific training examples for each interaction class. In this paper, we address both challenges and focus on the latter. Specifically, we introduce a method to train body part detectors from nonspecific images with pose information. Such resources are widely available. We introduce a training scheme and an adapted DPM formulation to allow for the inclusion of this auxiliary data. We perform cross-dataset experiments to evaluate the generalization performance of our method. We demonstrate that our method can still achieve decent performance, from as few as five training examples.

Knowledge base for a compositional expert system consists of a set of IF THEN rules with uncertainties expressed as weights. During consultation for a particular case, all aplicable rules are combined and weighhs of goals (final diagnoses or recommendations) are computed. The main problem when eliciting such knowledge base from an expert is the question of “correct” weights of rules.

Feature selection is a difficult but important issue in the field of machine learning and pattern recognition. In this paper, features for Chinese character recognition are selected by using inductive learning algorithms. The existing inductive learning method based on extension matrix requires precise consistency between positive example and negative example sets, which is very difficult to maintain in most practical cases. The traditional decision tree algorithm ID3 considers only the performance of the discriminating power while selecting features. However, in actual practice the consideration of the associated cost of feature extraction may become a significant concern. In addressing these problems we propose a modified extension matrix approach to select feature subset from the training example set with noises. A decision tree algorithm based on information gain and cost evaluation is also proposed to facilitate cost consideration. The comparative experiments show that the proposed algorithms perform better than the existing inductive learning algorithms to a certain extent.

This paper proposes machine learning techniques to discover knowledge in a dataset in the form of if-then rules for the purpose of formulating queries for validation of a Bayesian belief network model of the same data. Although do-main expertise is often available, the query formulation task is tedious and laborious, and hence automation of query formulation is desirable. In an effort to automate the query formulation process, a machine learning algorithm is lev-eraged to discover knowledge in the form of if-then rules in the data from which the Bayesian belief network model under validation was also induced. The set of if-then rules are processed and filtered through domain expertise to identify a subset that consists of “interesting” and “significant” rules. The subset of interesting and significant rules is formulated into corresponding queries to be posed, for validation purposes, to the Bayesian belief network induced from the same dataset. The promise of the proposed methodology was assessed through an empirical study performed on a real-life dataset, the National Crime Victimization Survey, which has over 250 attributes and well over 200,000 data points. The study demonstrated that the proposed approach is feasible and provides automation, in part, of the query formulation process for validation of a complex probabilistic model, which culminates in substantial savings for the need for human expert involvement and investment.

Support Vector Machines (SVMs) are a relatively new generation of techniques for classification and regression problems. These are based on Statistical Learning Theory having its origins in Machine Learning, which is defined by Kohavi and Foster (1998) as, ...Machine Learning is the field of scientific study that concentrates on induction algorithms and on other algorithms that can be said to “learn.”

A hyperbox classifier model for identifying secure carbon dioxide reservoirs

The paper examines application of a machine learning technique, decision tree induction, to the development of CBR systems. The decision tree generated by the induction algorithm is generally used to create a set of rules for a traditional rule based system; however, it may be used to determine the weight (i.e., importance) of features for classification. This approach combines the flexibility of CBR systems with the machine learning ability to automatically derive decision criteria. We experimentally assess the tolerance of rule induction to noise and missing values and find it surprisingly resilient to poor quality training data. This bodes well for use of rule induction to allow the process of CBR system development to be more independent of expert judgement.

A very important issue faced by researchers and practitioners who use industrial and research databases is incompleteness of data, usually in terms of missing or erroneous values. While some of data analysis algorithms can work with incomplete data, a large portion of them require complete data. Therefore, different strategies, such as deletion of incomplete examples, and imputation (filling) of missing values through variety of statistical and machine learning (ML) procedures, are developed to preprocess the incomplete data. This study concentrates on performing experimental analysis of several algorithms for imputation of missing values, which range from simple statistical algorithms like mean and hot deck imputation to imputation algorithms that work based on application of inductive ML algorithms. Three major families of ML algorithms, such as probabilistic algorithms (e.g. Naive Bayes), decision tree algorithms (e.g. C4.5), and decision rule algorithms (e.g. CLIP4), are used to implement the ML based imputation algorithms. The analysis is carried out using a comprehensive range of databases, for which missing values were introduced randomly. The goal of this paper is to provide general guidelines on selection of suitable data imputation algorithms based on characteristics of the data. The guidelines are developed by performing a comprehensive experimental comparison of performance of different data imputation algorithms.

Inductive learning algorithms typically use a set of labeled examples to learn class descriptions for a set of user-specified concepts of interest. In practice, labeling the training examples is a tedious, time consuming, error-prone process. Furthermore, in some applications, the labeling of each example also may be extremely expensive (e.g., it may require running costly laboratory tests). In order to reduce the number of labeled examples that are required for learning the concepts of interest, researchers proposed a variety of methods, such as active learning, semi-supervised learning, and meta-learning.

EXEMPLARY PROGRAMMING IN RITA

Maintaining proper operation of adaptive protection schemes is one of the main challenges that must be considered for smart grid deployment. The use of reliable cyber detection and protection systems boosts the preparedness potential of the network as required by National Infrastructure Protection Plans (NIPPS). In an effort to enhance grid cyber-physical resilience, this paper proposes a tool to enable attack detection in protective relays to tackle the problem of compromising their online settings by cyber attackers. Implementing the tool first involves an offline phase in which Monte Carlo simulation is used to generate a training dataset. Using rough set classification, a set of If-Then rules is obtained for each relay and loaded to the relays at the initialization stage. The second phase occurs during online operation, with each updated setting checked by the corresponding relay’s built-in tool to determine whether the settings received are genuine or compromised. A test dataset was generated to assess tool performance using the modified IEEE 34-bus test feeder. Several assessment measures have been used for performance evaluation and their results demonstrate the tool’s superior ability to classify settings efficiently using physical properties only.