Regional Institute of Management and Technology

Mandi Gobindgarh, India

Regional Institute of Management and Technology

Mandi Gobindgarh, India
SEARCH FILTERS
Time filter
Source Type

Dhaliwal D.S.,RIMIT Institute of Engineering and Technology | Sandhu P.S.,Rayat and Bahra Institute of Engineering and Bio Technology | Panda S.N.,Regional Institute of Management and Technology
World Academy of Science, Engineering and Technology | Year: 2011

In this paper, an enhancement of the k-Nearest Neighbor (k-NN) algorithm is proposed by incorporating min-max normalization of data as the initial stage before classification via the conventional k-NN algorithm and outlier removal as the final step. Under the proposed method, raw data is first normalized and the outlyingness factor (Oi) for each observation computed. A threshold value of 1.0 is used to detect outliers and observations with O1<1.0 fed for clustering by the k-NN method. In this study, the training set consisted of biological data derived from the Munich Information Center for Protein Sequences (MIPS) database. The algorithm was implemented in the PHP: hypertext preprocessor (PHP) scripting language. The data used was stored in a database implemented using MySQL using the Windows platform. The user interface for the application was constructed using advanced html using the Notepad text editor and linked to the backend using the PHP language. Using a dataset of 200 observations and a K parameter of 10, the outcomes obtained via the enhanced method were compared to those obtained via the conventional k-NN method. Comparisons of the results were made using the rand index. Results indicate that, compared to the naïve k-NN method, the enhanced method returns significantly improved performance.


Sandhu P.S.,Rayat and Bahra Institute of Engineering and Bio Technology | Dhaliwal D.S.,RIMIT Institute of Engineering and Technology | Panda S.N.,Regional Institute of Management and Technology
International Journal of Physical Sciences | Year: 2011

Association rule mining has been an area of active research in the field of knowledge discovery and numerous algorithms have been developed to this end. Of late, data mining researchers have improved upon the quality of association rule mining for business development by incorporating the influential factors like value (utility), quantity of items sold (weight) and more, for the mining of association patterns. In this paper, we propose an efficient approach based on weight factor and utility for effectual mining of significant association rules. Initially, the proposed approach makes use of the traditional Apriori algorithm to generate a set of association rules from a database. The proposed approach exploits the anti-monotone property of the Apriori algorithm, which states that for a k-itemset to be frequent all (k-1) subsets of this itemset also have to be frequent. Subsequently, the set of association rules mined are subjected to weightage (W-gain) and utility (U-gain) constraints, and for every association rule mined, a combined utility weighted score (UW-Score) is computed. Ultimately, we determine a subset of valuable association rules based on the UW-Score computed. The experimental results demonstrate the effectiveness of the proposed approach in generating high utility association rules that can be lucratively applied for business development. © 2011 Academic Journals.


Randhawa R.,Regional Institute of Management and Technology | Kaler R.S.,Thapar University
Optik | Year: 2010

In this paper, the impacts of Polarization Mode Dispersion (PMD) on the performance of high-speed optical communication system have been reported at different bit rates. The two systems are modeled using older fibers with same PMD coefficient at different bit rates and third is with the new fiber with less PMD coefficient than that of the previous two. The attenuation, chromatic dispersion and non-linear effects have been disabled, so that all the variation of the results is due to PMD. The bit rate is varied from 2.5 to 40 Gbps and the length is varied from 1000 to 20,000 km. It is shown that the impact of PMD increases with the bit rate of system. It is also reported that the impact of PMD becomes intolerable at the bit rates of more than 40 Gbps. And also the PMD produces very minute impact on the system performance for same bit rate with the variation in the fiber length. © 2009 Elsevier GmbH. All rights reserved.


Kaler R.,Regional Institute of Management and Technology | Kaler R.S.,Thapar University
Optik | Year: 2010

In this paper, the dispersion maps are compared in the presence of fiber nonlinearities in 10 and 40 Gbps carrier suppressed return to zero (CS-RZ) systems. It is observed that the hybrid-compensation is the best dispersion mapping technique, which reduces the BER produced due to the fiber nonlinearities to the more extent than that of pre-compensation and the post-compensation. This technique increases the repeater length almost up to double than that of pre- and post-compensation. The pre-compensation shows the highest BER followed by post-compensation. It is also observed that with increase in the input bit rate, the BER for all three compensation techniques is also increased. But the still the hybrid-compensation reduce the BER up to more extent. The BER also increases with the increase in input optical power due to nonlinear impairments on the link at high powers. © 2009 Elsevier GmbH. All rights reserved.


Sandhu P.S.,Rayat And Bahra Institute Of Engg And Bio Technology | Dhaliwal D.S.,Rimit Institute Of Engg And Technology | Panda S.N.,Regional Institute of Management and Technology | Bisht A.,Rayat And Bahra Institute Of Engg And Bio Technology
2nd International Conference on Computer and Network Technology, ICCNT 2010 | Year: 2010

Association rule mining has been an area of active research in the field of knowledge discovery and numerous algorithms have been developed to this end. Of late, data mining researchers have improved upon the quality of association rule mining for business development by incorporating the influential factors like value (utility), quantity of items sold (weight) and more, for the mining of association patterns. In this paper, we propose an efficient approach based on weight factor and utility for effectual mining of significant association rules. Initially, the proposed approach makes use of the traditional Apriori algorithm to generate a set of association rules from a database. The proposed approach exploits the anti-monotone property of the Apriori algorithm, which states that for a k-itemset to be frequent all (k-1) subsets of this itemset also have to be frequent. Subsequently, the set of association rules mined are subjected to weightage (W-gain) and utility (U-gain) constraints, and for every association rule mined, a combined Utility Weighted Score (UW-Score) is computed. Ultimately, we determine a subset of valuable association rules based on the UW-Score computed. The experimental results demonstrate the effectiveness of the proposed approach in generating high utility association rules that can be lucratively applied for business development. © 2010 IEEE.


Kaler R.,Regional Institute of Management and Technology | Kaler R.S.,Thapar University
Optik | Year: 2011

In this paper, we have demonstrated the quality-of-service offered by the metropolitan area network which is based on optical cross connect (OXC) and arrayed waveguide grating (AWG) demultiplexer operating at 10 Gb/s with 0.1 nm channel spacing for NRZ signal transmission. The data is successfully transmitted to a distance of 40 km with a reasonably good BER of 2.388 × 10-35. The OXC and AWG demultiplexers in the proposed architecture allow incremental expansion in terms of the number of wavelength channels to be transmitted. Dispersion and crosstalk are the main signal-degrading factors arising from the operation of the OXC and the effectiveness of each factor is individually investigated. © 2010 Elsevier GmbH. All rights reserved.


Kaler R.,Regional Institute of Management and Technology | Kaler R.S.,Thapar University
Optik | Year: 2011

In this paper, we have analyzed the performance and feasibility for the metropolitan area network based on arrayed waveguide grating (AWG) multiplexers and arrayed waveguide grating (AWG) demultiplexers operating at the bitrate of 10 Gb/s. In the network, the data is successfully transmitted to a distance of 50 km with a very low BER of 1 × 10-40 thus improving the performance over AWG star based networks. Here, we have observed that arrayed waveguide gratings based multiplexers and demultiplexers for WDM applications prove to be capable of precise multiplexing and demultiplexing of a large number of channels with relatively low losses. This paper also presents the comparative investigation and suitability of various data formats like NRZ Rectangular, NRZ Raised cosine, RZ Rectangular, RZ Raised cosine and RZ super Gaussian for optical transmission link. It has been shown that RZ Raised cosine yields the highest value of Q, good eye opening and lowest BER. © 2010 Elsevier GmbH. All rights reserved.


Kundra P.,Punjab Technical University | Sadawarti H.,Regional Institute of Management and Technology
Research Journal of Information Technology | Year: 2015

Swarm intelligence is a global research area to improve the optimization of various soft computing and nature inspired techniques. In this study, we have applied the hybrid algorithm of Cuckoo Search (CS) and Particle Swarm Optimization (PSO) for remote sensing image classification of natural terrain features. Remote sensing is the method of acquiring, processing and interpreting the satellite images and related geo-spatial data without any physical contact of that region. The main advantage of using the hybrid concept is that the search strategy used in CS for finding the best host nest for cuckoo egg is resolved by the best position of PSO concept. By using this proposed algorithm, it becomes easier to classify the terrain features and obtained results shows the higher efficiency and greater kappa coefficient value as compare to other swarm intelligence techniques. We have successfully applied the hybridization of Cuckoo Search (CS) and Particle Swarm Optimization (PSO) for classifying diversified land cover areas in a remote sensing satellite image. © 2015 Academic Journals Inc.


Kaler R.,Regional Institute of Management and Technology | Kaler R.S.,Thapar University
Optik | Year: 2011

In this paper, we compare the Gain and Noise figure characteristics of physical EDFAs and Compact EDFAs in an optical system consisting of cascade of both the amplifiers. We demonstrate the gain, noise figure variations of a forward pumped EDFA and Compact EDFAs as functions of Er3+ fiber length, injected pump power and up-conversion co-efficient. It is observed that the gain becomes constant when the length of both the amplifiers reaches above 20 m. The comparison shows that the higher gain with flatter output is obtained in case of Compact EDFAs than Physical EDFAs in a system consisting of chain of both the amplifiers. It is further investigated that the agreement between the Compact and Physical EDFA models is good up to 10 m with the no up-conversion co-efficient. Also, the noise figure obtained in case of Physical EDFA is higher than Compact EDFAs when same amplifier length is more than 20 m and then becomes constant for both the amplifiers. © 2010 Elsevier GmbH. All rights reserved.

Loading Regional Institute of Management and Technology collaborators
Loading Regional Institute of Management and Technology collaborators