Search results

Mobile phone-based human activity recognition (HAR) consists of inferring user’s activity type from the analysis of the inertial mobile sensor data. This paper aims to mainly introduce a new classification approach called adaptive k-nearest neighbors (AKNN) for intelligent HAR using smartphone inertial sensors with a potential real-time implementation on smartphone platform.

The proposed method puts forward several modification on AKNN baseline by using kernel discriminant analysis for feature reduction and hybridizing weighted support vector machines and KNN to tackle imbalanced class data set.

Extensive experiments on a five large scale daily activity recognition data set have been performed to demonstrate the effectiveness of the method in terms of error rate, recall, precision, F1-score and computational/memory resources, with several comparison with state-of-the art methods and other hybridization modes. The results showed that the proposed method can achieve more than 50% improvement in error rate metric and up to 5.6% in F1-score. The training phase is also shown to be reduced by a factor of six compared to baseline, which provides solid assets for smartphone implementation.

This work builds a bridge to already growing work in machine learning related to learning with small data set. Besides, the availability of systems that are able to perform on flight activity recognition on smartphone will have a significant impact in the field of pervasive health care, supporting a variety of practical applications such as elderly care, ambient assisted living and remote monitoring.

The purpose of this study is to build and test an accurate offline model by using only a compact training data that can reduce the computational and memory complexity of the system. This provides grounds for developing new innovative hybridization modes in the context of daily activity recognition and smartphone-based implementation. This study demonstrates that the new AKNN is able to classify the data without any training step because it does not use any model for fitting and only uses memory resources to store the corresponding support vectors.

The paper aims to answer the question “can the bipolar Negative‐Neutral‐Positive logic be extended and motivated in some probabilistic framework?”

Using the context of cognitive map interpretation of the conjunction and disjunction connectives in bipolar logic, three probabilistic causal reasoning have been put forward. The first one is based on the infinitesimal representation of material implication while the second one relies on the qualitative representation developed by Suppes and Cartwright. In both cases special conditions for transitivity of inference and multiple inputs scenarios are examined. The third developed approach implicitly omits the cognitive interpretation and rather relies on the idea that the causal independence structure can be substituted by some functional that combines independent inputs in such a way to force the output to be in full agreement with results expected through the conjunctive and disjunctive connectives.

The paper reports several theoretical findings regarding the different conditions ensuring the agreement and equivalence between the bipolar logic connectives and their probabilistic counterparts for each proposal. The paper also provides useful insights to link the finding to probabilistic argumentation system where pro and con arguments are considered simultaneously.

The paper offers theoretical basis for researchers investigating different categories of logics and contribute to the discussion linking logic to probability.

The task of identifying activity classes from sensor information in smart home is very challenging because of the imbalanced nature of such data set where some activities occur more frequently than others. Typically probabilistic models such as Hidden Markov Model (HMM) and Conditional Random Fields (CRF) are known as commonly employed for such purpose. The paper aims to discuss these issues.

In this work, the authors propose a robust strategy combining the Synthetic Minority Over-sampling Technique (SMOTE) with Cost Sensitive Support Vector Machines (CS-SVM) with an adaptive tuning of cost parameter in order to handle imbalanced data problem.

The results have demonstrated the usefulness of the approach through comparison with state of art of approaches including HMM, CRF, the traditional C-Support vector machines (C-SVM) and the Cost-Sensitive-SVM (CS-SVM) for classifying the activities using binary and ubiquitous sensors.

Performance metrics in the experiment/simulation include Accuracy, Precision/Recall and F measure.

In previous work, the algebraical properties of this rule and its relationship with other generalized operator was studied. In this paper, the aim is to focus on one of the previous steps, which consists in certainty qualification, and it is investigated how this factor influences the behavior of the induced combination rule.

Dubois and Prade have proposed an adaptive combination rule that moves gradually from a conjunctive mode to a disjunctive mode as soon as the conflict between the sources increases. The proposal can be viewed as a result of some rational steps. This includes: conjunctive combination; re‐normalization of a subnormal result that may results from conjunctive operation where the lack of normalization is interpreted as a conflict; certainty qualification; restriction of the conflict influence; generalization to more than two sources.

Algebraical properties of the proposals have been investigated and illustrations of some special cases are highlighted and evaluated. Further studies continue in Part II.

New functional adaptive rules are put forward based on Residual implicators and t‐conorm operators.

The first part of this issue investigated the properties of the adaptive rule initially proposed by Dubois and Prade given in the framework of possibility theory, when the certainty qualification is rather expressed in more general t‐norms and t‐conorms connectives. This led to two new family of adaptive rules expressed using residual implication and t‐conorm connective, respectively. The problem of addressing uncertain inputs has also been examined and a waved decomposition has been proposed in PII we study adaptative combinations with incomplete certainty qualification. However, another problem that arises when combining uncertain inputs consists of the relationship between the certainty attached to the inputs and the certainty attached to the output, conceptualized by the resulting distribution when using adaptive combination rule. In other words, how does the combination rule improves or deteriorates the certainty of the overall system? This paper seeks to address this issue.

This paper fully addresses this issue and attempts to evaluate the combination rule from the certainty viewpoint attached to the result in comparison to initial certainty values attached to the inputs.

Especially, it has been proven that under certain hypotheses, the rule allows the user to hide the local certainties attached to the initial inputs, while highlighting only the certainty due to the lack of consistency among the sources.

New functional adaptative rules are put forward based on residual implicators and t‐conorm operators.

To give a mathematical expression of what could be called the internal time of a dynamical system, a time which is different from the external or reference time.

The paper introduces a general mathematical definition of internal duration and so of internal time. Then we consider the case of an explosion followed by an implosion, which we apply to cosmology and physiology. The case of diffusion is also presented.

The internal time is generally different from the reference time. In certain cases to a finite reference duration may correspond an infinite internal duration.

Our formulations may help to understand certain aspects of cosmology, physiology and more generally of the evolution of dynamical systems.

For example, the physiology of ageing.

The consideration of the square of the speed of evolution, at instant t, of a dynamical system for measuring the internal duration of interval (t,t+dt) is original, as well as its consequences.