Abstract

Regression algorithms are used for prediction (including forecasting of time-series data), inference, hypothesis testing, and modeling of causal relationships. Statistical approaches although popular, are are not generic in that they require the user to make an intelligent guess about the form of the regression equation. In this paper we present a new regression algorithm GEAR – Generic, Efficient, Accurate kNN-based Regression. In addition to this, GEAR is simple and outlier-resilient. These desirable features make GEAR a very attractive alternative to existing approaches. Our experimental study compares GEAR with fourteen other algorithms on five standard real datasets, and shows that GEAR is more accurate than all its competitors.

Abstract

This paper presents Perturbed Frequent Itemset based Classification Technique (PERFICT),a novel associative classification approach based on perturbed frequent itemsets. Most of the existing associative classifiers work well on transactional data where each record contains a set of boolean items.They are not very effective in general for relational data that typically contains real valued attributes.In PERFICT, we handle real attributes by treating items as (attribute,value) pairs, where the value is not the original one, but is perturbed by a small amount and is a range based value. We also propose our own similarity measure which captures the nature of real valued attributes and provide effective weights for the itemsets. The probabilistic contributions of different itemsets is taken into considerations during classification. Some of the applications where such a technique is useful are in signal classification, medical diagnosis and handwriting recognition. Experiments conducted on the UCI Repository datasets show that PERFICT is highly competitive in terms of accuracy in comparison with popular associative classification methods.

Abstract

Uncertainty is inherently present in many real-world domains like images. Analyses of such uncertain data using traditional certain-data-oriented techniques do not achieve best possible accuracy. UACI introduces the concept of representing images in the form of a probabilistic or uncertain model using interest points in images. This model is an uncertain-data-based adaptation of Bag of Words, with each image not only represented by the visual words that it contains, but also their respective probabilities of occurrence in the image. UACI uses an Associative Classification approach to leverage latent frequent patterns in images for the identification of object classes. Unlike most image classifiers, which rely on positive and negative class sets (generally very vague) for training, UACI uses only positive class images for training. We empirically compare UACI with three other state-of-the-art image classifiers, and show that UACI performs much better than the other classifying approaches.

Abstract

In the present study, we have constructed and analysed the metabolic networks of virulent and non-virulent strain of E. coli. Concepts from graph theory are used to systematically determine how differences in basic metabolism of different strains of E. coliare reflected at the systems level. Analysis of metabolic networks can help us to understand the evolutionary history of cellular metabolic process and we can also identify the potential drug target by comparing the networks.

Abstract

Two classes of networks that have been extensively studied in the analysis of physical systems are: (i) regular networks, wherein each node interacts with a specified number of neighboring nodes on geometrical lattices, and (ii) random networks, wherein every pair of nodes have a fixed probability of interacting with each other. Recently much attention is being focused on a class of network models that are neither strictly regular nor completely random, but are somewhere in-between and exhibit properties of both, called small world networks. These networks are shown to exhibit high clustering (i.e., nodes sharing a common neighbor have a higher probability of being connected to each other than to other nodes) and a low average path length (the average of the shortest distance/path between every pair of nodes in the network). Examples for small-world networks are found to occur widely across the biological (e.g., neural connection patterns), social (e.g., friendship network, co-authorship) and technological (e.g., the world wide web) domains. Also it has been observed that a large number of real complex systems exhibit power-law behaviour in their degree distribution, i.e., a few nodes have a very high degree. Such networks are referred to as scale-free networks. Thus, non-standard topologies with long-range connections (i.e., non-local diffusion) and uneven degree distributions are not uncommon in real-life systems and may provide different kinds of spatiotemporal dynamics depending on the extent of non-local diffusion. Here we discuss the characterization and control of spatiotemporal dynamics on four different network topologies, viz., (i) regular, (ii) random, (iii) small-world, and (iv) scale-free by external perturbation or pinning a few nodes in the network. This would provide us insight into the role of the network topology (the underlying connectivity structure) on the dynamics of the systems defined on such networks and the efficacy with which the dynamics can be the controlled or manipulated. An extensively studied example of a nonlinear system exhibiting a wide variety of complex dynamics ranging from simple periodic behavior to chaos is the logistic map. Here we define coupled logistic maps on the four different topologies and systematically investigate the control by external perturbation/pinning for two chaotic regimes: (i) r = 3.6 (weak-chaos), and (ii) r = 3.9 (strong chaos). Our preliminary results show that pinning nodes at regularly spaced intervals, 2nd, 4th, etc., the pinning strength required on a small-world network is similar to that in case of regular networks. For 25% of the nodes pinned, the dynamics on the scale-free topology is controllable but on random networks, it is not. However, on pinning nodes having high centrality measures, viz., degree, betweenness and closeness, we observe that control of the spatiotemporal dynamics is achieved, by pinning only 10% of high degree/ betweenness nodes in low chaotic regime on both small-world and scale-free networks. Complete control of the network is not observed on pinning nodes with high closeness values. For strongly chaotic dynamics, control is achievable only on scale-free networks on pinning 20% of nodes having either high degree, betweenness, or closeness values. Key Words: Dynamics, Control, Topologies, Coupled Logistic Maps, Chaos, Regular, Smallworld, Random, Scale-free networks, Centrality.

Abstract

In this study we analyze the topology and structure of Airport Network of India (ANI) using the graph theoretic approach. We investigate the transmission flow in the network using graph centrality measures. We observe that though ANI exhibits scale-free behaviour, it differs from Barabasi-Albert scale-free model and we suggest a scale-free model that results in high clustering coefficient to be most suitable to model ANI. The analysis is not only useful in planning the infrastructure and expansion of the air-traffic connectivity, but also in managing the flow of transportation during emergencies such as accidental failure of the airport, closing down the airport due to unexpected climate changes, terrorist attacks etc. It has been observed that densely connected air-transportation networks play a major role in the spread of infectious diseases, viz., Avian-influenza, Swine-Flu, etc., turning from epidemic into pandemic. Knowledge of the connectivity pattern for reduction of flights on certain routes can help to reduce the spread of the disease.

Abstract

For speaker recognition studies, it is necessary to process the speech signal suitably to capture the speaker-specific infor-mation. There is complementary speaker-specific information in the excitation source and vocal tract system characteristics.Therefore it is necessary to separate these components, even approximately, from the speech signal. We propose linear pre-diction (LP) residual and LP coefficients to represent these two components. Analysis is performed in a pitch synchronous manner in order to focus on the significant portion of the speech signal in each glottal cycle, and also to reduce the artifacts of digital signal processing on the extracted features. Finally, the speaker-specific information is captured from the excitation and the vocal tract system components using auto associative neural networks (AANN) models. We show that the pitch synchronous extraction of information from the residual and vocal tract sys-tem bring out the speaker-specific information much better than using the pitch asynchronous analysis as in the traditional block processing using an analysis window of fixed size

Abstract

The voice onset time (VOT) combines the temporal and frequency structure over very short duration. This makes the VOT detection task difficult. But the VOT is an important temporal feature. In this paper we propose a new method for the detection of VOT in speech utterances. The method uses Fourier-Bessel (FB) expansion followed by amplitude and frequency modulated (AM-FM) signal model. The FB expansion is used to emphasize the vowel and consonant regions of stop consonant vowel (SCV) units (/ka/, /ta/, and /pa/). The emphasized signals using selected range of FB coefficients have been considered narrow-band AM-FM signals. The discrete energy separation algorithm (DESA) has been used to estimate amplitude envelope (AE) function of the AM-FM model. The estimated AE function is explored for the detection of VOT. The VOT detection studies are carried out using the speech corpus consists of utterances of various male and female speakers.

Abstract

Glottal closure instants (GCI) are the instants of significant excitation of the vocal tract system during the production of speech. The performance of many speech analysis and synthesis approaches depends on accurate estimation of GCIs. A new method for detection of GCI locations in speech utterances using Fourier-Bessel (FB) expansion and amplitude and frequency modulated (AM-FM) signal model is proposed in this paper. The inherent filtering property of the FB expansion is used to weaken the effect of formants in the speech utterances. The band-limited signal reconstructed using FB coefficients has been considered as an AM-FM signal. The discrete-energy separation algorithm (DESA) has been used to estimate the amplitude envelope (AE) function of the AM-FM signal model. The estimated AE function is explored for detection of GCIs. The CMU-Arctic database is used in this work for the validation of the proposed method for GCI detection with the electro-glottograph (EGG) as a reference. It has been observed that the detected GCI by the proposed method seems to provide accurate location of GCIs.

Abstract

Sensor networks consist of small motes attached with sensors to measure ambient parameters like temperature, humidity and light. As these motes are unreliable due to wireless link quality and also the data measuring sensors cannot be calibrated accurately for a given applications need. The unique data fusion needs are that parameter being measured is distributed across the network and needs to be computed reliably and with minimum overhead and redundancy due to data value being correlated. We show the asymptotic complexity of topology control when applied to power-aware routing is scalable and argue that the accuracy and reliability of the estimated sensor values can be accurately predicted for the physical value being sensed and aggregating. A prefix-based routing protocol is used for data-centric storage, which allows querying distributed parameters using a KEY, VALUE pairs without the need of the sensor node to know its exact geographic information. Intelligent sensor information processing, which is driven by these requirements, is discussed under the framework INSPIRE-DB.