Muhammad Awais Azam

Novel centroid selection approaches for KMeans-clustering based recommender systems

Recommender systems have the ability to filter unseen information for predicting whether a particular user would prefer a given item when making a choice. Over the years, this process has been dependent on robust applications of data mining and machine learning techniques, which are known to have scalability issues when being applied for recommender systems. In this paper, we propose a k-means clustering-based recommendation algorithm, which addresses the scalability issues associated with traditional recommender systems. An issue with traditional k-means clustering algorithms is that they choose the initial k centroid randomly, which leads to inaccurate recommendations and increased cost for offline training of clusters. The work in this paper highlights how centroid selection in k-means based recommender systems can improve performance as well as being cost saving. The proposed centroid selection method has the ability to exploit underlying data correlation structures, which has been proven to exhibit superior accuracy and performance in comparison to the traditional centroid selection strategies, which choose centroids randomly. The proposed approach has been validated with an extensive set of experiments based on five different datasets (from movies, books, and music domain). These experiments prove that the proposed approach provides a better quality cluster and converges quicker than existing approaches, which in turn improves accuracy of the recommendation provided.

Frequent Pattern Mining Algorithms for Finding Associated Frequent Patterns for Data Streams: A Survey☆

Abstract Pattern recognition is seen as a major challenge within the field of data mining and knowledge discovery. For the work in this paper, we have analyzed a range of widely used algorithms for finding frequent patterns with the purpose of discovering how these algorithms can be used to obtain frequent patterns over large transactional databases. This has been presented in the form of a comparative study of the following algorithms: Apriori algorithm, Frequent Pattern (FP) Growth algorithm, Rapid Association Rule Mining (RARM), ECLAT algorithm and Associated Sensor Pattern Mining of Data Stream (ASPMS) frequent pattern mining algorithms. This study also focuses on each of the algorithms strengths and weaknesses for finding patterns among large item sets in database systems.

Improved local spectrum sensing for cognitive radio networks

The successful deployment of dynamic spectrum access requires cognitive radio (CR) to more accurately find the unoccupied portion of the spectrum. An accurate spectrum sensing technique can reduce the probability of false alarms and misdetection. Cooperative spectrum sensing is usually employed to achieve accuracy and improve reliability, but at the cost of cooperation overhead among CR users. This overhead can be reduced by improving local spectrum sensing accuracy. Several signal processing techniques for transmitter detection have been proposed in the literature but more sophisticated approaches are needed to enhance sensing efficiency. This article proposes a two-stage local spectrum sensing approach. In the first stage, each CR performs existing spectrum sensing techniques, i.e., energy detection, matched filter detection, and cyclostationary detection. In the second stage, the output from each technique is combined using fuzzy logic in order to deduce the presence or absence of a primary transmitter. Simulation results verify that our proposed technique outperforms existing local spectrum sensing techniques. The proposed approach shows significant improvement in sensing accuracy by exhibiting a higher probability of detection and low false alarms. The mean detection time of the proposed scheme is equivalent to that of cyclostationary detection.

Fully Distributed Cooperative Spectrum Sensing for Cognitive Radio Ad Hoc Networks

Cognitive radios are indispensable to shift from conventional spectrum assignment to dynamic spectrum access. These are intelligent radios having the capability of sensing the radio environment and reconfiguring the operating parameters dynamically. Recently cognitive radios are considered for ad hoc environment. Cognitive radio ad hoc network have several unique features other than inherited features of ad hoc network like distributed multi-hop architecture, variant network topology and opportunistic spectrum availability. It is required for cognitive radio to find unutilized portion of the spectrum more accurately for successful deployment of cognitive radio ad hoc networks. In cooperative spectrum sensing, multiple cognitive radio users cooperate to detect the presence or absence of primary user to improve the detection performance. Ad hoc networks are dynamic in nature and have no central entity for data fusion. Therefore cognitive radio ad hoc networks need fully distributed cooperative spectrum sensing. In this paper, a gradient based fully distributed cooperative spectrum sensing is proposed for cognitive radio ad hoc networks. The proposed scheme is analyzed from the perspective of reliable sensing and energy consumption. Simulation results show the effectiveness of the proposed scheme.

Modeling of cutting speed (CS) for HSLA steel in wire electrical discharge machining (WEDM) using moly wire

Abstract The high capital costs of wire electrical discharge machining (WEDM) equipment necessitate optimal utilization of the WEDM process and equipment. Cutting speed (CS) is a key performance measure to achieve this objective. However, process parameters of WEDM greatly hamper CS and hence productivity and machining efficiency. It is therefore essential to pick the right combination of parameters to attain better CSs. In this paper, five process parameters which include pulse on-time, pulse off-time, pulse frequency, power, and wire speed were used to develop an empirical relationship between process parameters and CS. A regression model based on experimental data was developed and validated through confirmation tests. Experiments have been conducted on high-strength low-alloy steel using molybdenum wire. Analysis of variance was applied to segregate significant process parameters and it was revealed that pulse off-time, power, and pulse frequency were the major parameters affecting CS. Contour plots have been established to select the best process parameters in addition to the developed model. Stability of moly wire was also explored using scanning electron microscope and energy dispersive spectroscopy analysis. Results showed that moly wire retains its original surface quality and dimensions which contributes to dimensional accuracy of parts.

Tiered approach to infer the behaviour of low entropy mobile people

Being able to understand human behaviour and monitoring daily life activities is seen as a significant approach for alleviating functional decline among elderly people. The aim of the research work presented in this paper is to investigate a mechanism that can recognise high level activities and behaviour of low entropy people in order to help them improve their health related daily life activities by using wireless proximity data (e.g. Bluetooth, Wi-Fi). A number of scenarios and experiments are designed to prove the validity of the proposed methodology. Using wireless proximity data for activity recognition enhances the intrusion into personal privacy and helps exploiting important structures in human behaviour.

Chipless slot resonators for IoT system identification

In this paper chipless RFID tag with integrated sensor for IoT application is presented. The tag is capable of transmitting information of 9-bit data. The tag structure is analyzed for both FR4 and Kapton HN substrates having different dielectric constants. It has been observed that with the change in dielectric permittivity, there will be shift in resonances. The tag can be used for monitoring and sensing moisture. The tag comprises of 9 ring resonators. The overall radius of designed chipless tag is 7mm. The compact chipless RFID tag is optimized for radio frequency ranges from 5.1-11.4 GHz using FR4 substrate and from 5.8-12.5 GHz using Kapton HN substrate. The novelty of this design relies on flexible nature of tag. The presented tag is very cheap and can be deployed for various low cost sensing applications.

Propagation Analysis of Malware Families in Mobile P2P Networks

Viral propagation modelling acts as sandbox for testing intensity of malware, understand patterns adopted for malware propagation and consequently help device strategies for malware detection. Success of P2P networks has encouraged mobile vendors to offer P2P services on mobile networks. Handheld mobile devices though constrained in memory, power and processing resources are capable of using communication technologies like Bluetooth, MMS, SMS, Infrared and WLAN services. Such versatility has however exposed mobile devices to threats like mobile P2P malware. With the number of mobile phone malware escalating to an alarming figure of more than one thousand, it has become ever more important to analyze the affects of propagation of such malware in the wild that could subsequently act as the baseline for protection against such malware. This paper initially presents propagation analysis of generic mobile P2P malware categories and then provides a detailed analysis of propagation of real-world malware from three malware families accommodating around 100 well known mobile P2P malware. Paper is aimed at providing a much needed insight into propagation characteristics of mobile P2P malware like their propagation speed and battery depletion affect.

Cooperative Spectrum Sensing for Cognitive Radio Networks Application: Performance Analysis for Realistic Channel Conditions

Cognitive radio is a key technology to overcome spectrum scarcity by using spectrum opportunistically. It can be applied to maritime wireless networks to provide more bandwidth and reduce communication cost. Spectrum sensing is a primary issue to develop cognitive radio networks. There are few challenges for spectrum sensing in maritime networks which are different from terrestrial networks, for example, sea’s surface channel properties. High probability of detection is required to achieve better network performance. In this paper, centralized cooperative spectrum sensing schemes are compared for maritime cognitive radio networks. The simulation results show that exiting schemes are well suitable for lower sea states but fail for higher sea states and we need to devise some advanced algorithms for spectrum sensing in the maritime wireless networks.

Behavioural analysis of low entropy mobile people using contextual information

The profusion of wireless enabled mobile devices in daily life routine and advancement in pervasive computing has opened new horizons to analyse and model the contextual information. This contextual information (for example, proximity data and location information) can be very helpful in analysing the human behaviours. Wireless proximity data can provide important information about the behaviour and daily life routines of an individual. In this paper, we used Bluetooth proximity data to validate this concept by detecting repeated activity patterns and behaviour of low entropy mobile people by using n-gram and correlative matrix techniques. Primary purpose is to find out whether contextual information obtained from Bluetooth proximity data is useful for activities and behaviour detection of individuals. Repeated patterns found in Bluetooth proximity data can also show the long term routines such as, monthly or yearly patterns in an individuals daily life that can further help to analyse more complex and abnormal routines of human behaviour.