|
Sr/No/ |
Title |
Pdf |
|
1 |
Recommender Systems Based on User location
Priyadharsini S, Priyadharshini P
Abstract —Recommender systems are used to gather user’s
opinions and analyze information to provide suggestions to the
user. Location based recommendations are used to rate the
locations based on user’s experience. Both user locations and
item locations can be given as an input we have to identify the
nearest path to reach the destination. The main problem in
existing systems is not to consider the spatial properties of users
and items. In this paper we propose LARS (Location Aware
Recommender System) can be used to produce recommendations
to spatial items. Our system can recommend the item which is
nearer to the user’s query location. This recommends the
querying user with nearest item based on the spatial ratings of
the user and the item. This employs both user partitioning and
travel penalty techniques to consider user preferences and
travelling distances for the user. It can also make high quality
location based recommendation in an efficient manner. The user
can prefer the item which is nearer to the user query location. In
our system, we consider spatial properties of users and items to
produce recommendations with high quality.
|
Download
|
|
2 |
Cost Evaluation Model in an Isolated Distribution Network
V.Renu,Dr.S.Jeyadevi
Abstract—Growing environmental concerns and the reduction
of fossil fuel resources has led to the use of renewable energy
resources on the electric power sector. Among the variable
renewable resources, wind power is assumed to have the most
favorable technical and economical prospects. In recent years,
renewable generation by wind turbine have drawn great
attention due to the issues related to energy saving and carbon
emission reduction. The power supply must be reliable in the
electric power sector. So the cost evaluation is important for
the effective planning and operation of the power system.
Renewable energies are inexhaustible, clean and they can be
used in a decentralized way. It can be regarded as one of the
fundamental premises for building a sustainable global
society. In this work, multi-objective optimization problem
(fuel cost, carbon emission cost) is considered. For
demonstration purpose IEEE- 30 bus system is considered.
|
Download
|
|
3 |
Parkinson Disease Detection from EMG Signal using
Neural Network
V. Sutha JebaKumari, D. Shanthi, P. Meha
Abstract—TParkinson’s disease is the second most common
neuro degenerative disorder next to Alzheimer. In this paper
we use artificial neural network to predict the onset of tremor
in patients with parkinson disease. Tremor is the hallmark of
Parkinson’s disease and the presence of tremor in PD is a
fundamental feature used in the determination of disease onset
and progression. Data for the network was obtained from
EMG signals. These signals allow us to look directly into the
muscle and helps in the measurement of muscular performance.
The statistical features from EMG signals are extracted and
classified using Neural Network as they provide a good
performance in the diagnosis task.
|
Download
|
|
4 |
A Study on Mortar Made with Manufactured Sand
B.Jeyaprabha, G.Elangovan, P.Prakash
Abstract— In this experimental work, the effect of
elevated temperature on the mechanical properties of mortar
of ratio 1:3 prepared by Manufactured sand has been
investigated. Test samples were exposed to temperatures of
200°, 400°, 600 and 800°C at a high temperature
programmable muffle furnace for exactly three hour. Out of 6
specimens 3 samples were quenched in water for rapid
cooling and the other 3 were kept aside for normal cooling at
atmospheric temperature and subsequently, compressive
strength test was conducted on the specimens.Test results
indicated that the compressive strength of mortar decreased
as the exposure temperature increased. Water quenched
mortars showed lesser residual strength than air cooled
specimens.
|
Download
|
|
5 |
Determination of spectral signatures of end members
Using pixel purity index in hyperspectral images
G.Vijaya Kumar, G. Veera Senthil Kumar, S. Vasuki
Abstract-In the recent years, Hyper spectral Image HSI)
analysis is found more useful in remote sensing applications,
compared with other imaging techniques, because of its high
dimensionality. Despite many algorithms have been developed for
spectral un-mixing, it is found to be a challenging task even now.
In contrast to that high spectral resolution, the hyperspectral
sensors may not able to cover many unknown signal sources
which cannot be able to visual inspection. In order to take
account of that unknown signal sources, the concept of Virtual
Dimensionality(VD) is introduced in this paper, To determine VD
for hyperspectral images, the Eigen thresholding methods are
popularly used (i.e. Harsanyi–Farrand–Chang (HFC) and Noise-
Whitened HFC (NWHFC). These methods provide more reliable
estimates of VD. Principal Component Analysis (PCA) has been
adopted for dimensionality reduction. Endmember extraction
and determination of spectral signatures are done using Pixel
Purity Index (PPI) in this paper. Experiment is carried out with
AVIRIS data set (Cuprite-Reflectance). The proposed algorithm
is implemented in MATLAB.
|
Download
|
|
6 |
Impact of Active Network Management Schemes for Distribution
Network Planning incorporating multiple DGs
R.Sangeetha, S. Nagalakshmi
Abstract -In order to meet out the increasing power demand, a
new trend of generating power locally at distribution voltage
level called as Distributed Generation (DG) has been evolved. DG
allows collection of energy from many sources and gives lower
ecological impacts and improved protection of supply. Large
amounts of new capacity; mainly wind power is needed
connected to power system, to achieve their renewable target for
worldwide. To obtain the maximum DG capacity, Active network
management schemes (ANM) are used. The major ANM schemes
include coordinated voltage control (CVC) of OLTCs and voltage
regulators, compensator reactive power control, DG’s power
factor control (PFC) and energy curtailment (EC). In this paper,
the impact of multiple DG configurations on CVC scheme is
studied. The paper proposes a multi-period with Particle swarm
optimization (PSO) algorithm, for assessing the maximum DG
capacity under ANM schemes considering variability of demand
and generation profiles (multi-period scenarios). The allocation
of DG units should be properly determined with the
consideration of different planning incentives. Hence in this
paper, PSO algorithm is chosen for, obtain the maximum DG
capacity. Compare with other optimization methods it is more
efficient. The proposed work is tested on IEEE 33 bus radial
distributed system.
|
Download
|
|
7 |
An efficient architecture in clock gating using
Wide spread adapting technique
R.Kiruthika, T. Kavitha
Abstract-The Clock gating reduces dynamic power
dissipation in synchronous circuits. A clock gating cell
(CGC) is inserted, which covers an internal latch that filters
out potential glitches from a block called as the gating
function. It uses a matching factor forms, such as strong and
weak matching. Strong matching identifies matches that are
explicitly present in the factored forms, and weak matching
seeks matches that are implicit in the logic and thus they are
more difficult to discover in existing system. The problem in
existing system involves a flip-flop architecture with higher
power consumption, gate count and circuit size . Hence a
modified Boolean function technique known as Wide Spread
Adapting (WSA) clock gating technique is implemented. The
weak and strong matching are done by WSA technique since
they do not require separate algorithms as in existing
system. WSA technique helps to reduce the clocking signal
and the gate pattern. Thus the proposed method achieves
reduced clock gating which also reduces delay, power, gate
count and area.
|
Download
|
|
8 |
Medical Image And Information Retrieval Based On Multiple Features Ofan Image
S.Nirmala,A.Akileshwari,A.Ramya,S.Kavitha
Abstract-Image and information retrieval system have shown great potential in supporting decision making in medical and web application. In an image retrieval features plays a key role.In exiting method they retrieved the images only based on Color and Content based image retrieval system is not efficient for all the time. Because it is only based on the Meta data. Image and information retrieval is one of the most applicable image processing techniques, which has been used extensively. Feature extraction is one of the most important procedures used for interpretation and indexing images in image retrieval system. Effective storage, indexing and managing a large number of image collections is a critical challenge in computer system. There are many proposed methods to overcome these problems. However, the rate of accurate image retrieval and efficient of retrieval is still an interesting field of research. The results show that it is possible to improve a system for CCM(Color Cooccurance Matrix), GLCM(Gray Level Cooccurane Matrix), PCM(Principle Cooccurance Matrix), and gabor filter based image retrieval by using an appropriately defined multifeature fusion model, which takes careful consideration of structure and distribution of visual features.
|
Download
|
|
9 |
A Heuristic Approach for Traffic Pattern
Discovery System in MANETs
R.Shobana Devi, A.Deepajothi
Abstract-Many Conceptual namelessness
upgrading systems have been proposed focused
around parcel encryption to secure the
correspondence. Namelessness of portable
specially appointed systems (Manets).
Notwithstanding, in this paper, we demonstrate
that Manets are still powerless under latent
measurable activity investigation assaults. To
exhibit how to find the correspondence designs
without unscrambling the caught bundles, we
display a novel measurable movement design
disclosure framework (STARS). STARS meet
expectations latently to perform activity
investigation in light of factual attributes of
caught crude activity. STARS are fit for finding
the sources, the ends of the line, and the end-toend
correspondence relations. Exact studies
exhibit that STARS attains great precision in
revealing the covered up activity designs.
|
Download
|
|
10 |
Medical Image And Information Retrieval Based On Multiple Features Ofan Image
S.Nirmala,A.Akileshwari,A.Ramya,S.Kavitha
Abstract─Image and information retrieval system have shown great potential in supporting decision making in medical and web application. In an image retrieval features plays a key role.In exiting method they retrieved the images only based on Color and Content based image retrieval system is not efficient for all the time. Because it is only based on the Meta data. Image and information retrieval is one of the most applicable image processing techniques, which has been used extensively. Feature extraction is one of the most important procedures used for interpretation and indexing images in image retrieval system. Effective storage, indexing and managing a large number of image collections is a critical challenge in computer system. There are many proposed methods to overcome these problems. However, the rate of accurate image retrieval and efficient of retrieval is still an interesting field of research. The results show that it is possible to improve a system for CCM(Color Cooccurance Matrix), GLCM(Gray Level Cooccurane Matrix), PCM(Principle Cooccurance Matrix), and gabor filter based image retrieval by using an appropriately defined multifeature fusion model, which takes careful consideration of structure and distribution of visual features.
|
Download
|