Published in 2015
S.Saranya, B. Vasumathi, M.Sakthivel
Load forecasting is an important component for power system energy management system. Forecasting
means estimating active loads at various load buses ahead of actual load occurrence Training data is classified using
Fuzzy Set Based Classification Method. Temperature data is classified into five fuzzy sets (Very Cold, Cold, Normal,
Hot and Very Hot). Relative Humidity is classified into four fuzzy sets (Very Dry, Dry, Humid and Very Humid). Day
Type is classified into four fuzzy sets (Post-Holiday, Weekday, Pre-Holiday and Holiday). So, depending upon the
temperature, relative humidity and day type, data is classified into eighty classes. After the classification, the neural
network is trained for various classes using the historical data. The multilayer neural network structure has been
used and the training is imparted using back propagation algorithm. . In this article, a Fuzzy Set Classified Neural
Network Approach for Short Term Load Forecasting is attempted and implemented using Matlab 6.5.
Fuzzy Set Based Classification, Training of Neural Network, Short term load forecasting
S. Lavanya, Dr. S. Palaniswami, G. Kavinraj
In medical field, class imbalance is one in every of the powerful issues in class of data mining, usually unbalanced
dataset found in varieties of applications and notably in medical centre caused from few styles of issues like uncertainty,
absence of information, imbalance, volumetric, misclassified value and degrades the performance of information mining in
terms of accuracy, cost, decision creating. K-Nearest Neighbor (KNN) and Cost Sensitive Learning algorithms with hybrid ways
used to solve the category imbalance disadvantage and to create superior result to existing solutions. Hospitalized datasets
square measure crazy imbalance quantitative relation is taken into thought from uci repository and mistreatment F-measure to
produce the accuracy, sensitive and confidence for patient data. And for future conceive to scale back the imbalance data issues.
In found multi-objective native search downside resolved mistreatment Pittsburgh model that is expensive and turn out rules for
little coverage.
K-Nearest Neighbor (KNN) , Cost Sensitive Learning, F-measure, Classifier Level Approaches, Data Level Approaches
R.Nallakumar, Dr. N. Sengottaiyan, R.Ramya Vinodini
Today the number of users in cloud computing are increasing tremendously due to its advantage of
providing flexible storage requirement. Due to the high volume and velocity of big data, it is an effective option to
store big data in the cloud Attribute-Based Encryption (ABE) is a promising technique to ensure the end-to-end
security of big data in the cloud. However, the policy updating has always been a challenging issue when ABE is used
to construct access control schemes. A trivial implementation is to let data owners retrieve the data and re-encrypt it
under the new access policy, and then send it back to the cloud. This method, however, incurs a high communication
overhead and heavy computation burden on data owners. In this project, we propose a novel scheme that enabling
efficient access control with dynamic policy updating for big data in the cloud. We focus on developing an
outsourced policy updating method for ABE systems. This method can avoid the over flow of data by reusing
previously encrypted data with old access policies and also we propose the automatic encryption of the data by
using a single algorithm so it reduces the work complexity of the data owners and time consumption in decryption
part. The analysis shows that our policy updating outsourcing scheme is correct, complete, secure and efficient.
Access control, Attribute Based Encryption (ABE), Policy Updating, Outsourcing, Big Data, Cloud.
T.Parameswaran,, Dr.C.PalaniSamy, G.Shenbagavalli
Cognitive radio is a emerging technology to solve the problems of spectrum inefficiency, and scarcity by providing the
vacant channel to the secondary users without disturbing the primary users. This paper presents Gymkhana, multi path route
discovery algorithm for cognitive radio mobile ad hoc networks. Data traffic congestion in the network can be avoided by
reducing the number of paths from source to destination. From the result of the routing algorithm trust value can be calculated.
In the proposed trust based CR-MANET routing protocol, the trust model involves two components: trust value from direct
observation and trust value from indirect observation. With direct observation from an observer node, the trust value calculated
using Bayesian inference, which is a type of uncertain reasoning when the full probability model can be defined. On the other
hand, with indirect observation, the trust value is obtained from neighbor nodes of the source of Secondary user, the trust value
is derived using the Dumpster-Shafer theory, which is another type of uncertain reasoning Combining these two trust values,
obtain more accurate trust values of the observed nodes in CR-MANETs
Cognitive radio Mobile ad hoc Networks, Channels, Trust Model, Packets
R.Nallakumar, Dr. N. Sengottaiyan, N.Babu
Cloud computing is arising as a prevalent data interactive paradigm to realize user’s data remotely stored in an
online cloud server. Cloud services provide great appropriateness for the users to enjoy the on-demand cloud
applications without considering the local underpinning limitations. During the data scare up, different users may be in a
collaborative relationship, and thus data sharing becomes indicative to achieve productive benefits. A cloud storage
system, consisting of a accumulate of storage servers, provides long-term storage services over the Internet. Storing
data in a third party’s cloud system spring serious concern over data acquaintance. General encryption schemes protect
data confidentiality, but also limit the component of the storage system because a few operations are supported over
encrypted data. Constructing a secure storage system that footing multiple functions is challenging when the storage
system is distributed and has no central authority. We propose vestibule proxy re-encryption scheme and integrate it
with a decentralized erasure code such that a secure distributed arcade system is formulated
cloud computing, proxy server, re-encryption, security.
R.Nallakumar, Dr. N. Sengottaiyan, S.Parthiparaj
Automation products have gotten magnified and wish of the automotive product additionally. Still the present
system face a lot of drawbacks in security within the current software package, order putting, producing the product in
step with the order and delivering the product. The foremost vital issue is computerizing all the information during a
centralized server and taking backups of previous records. So as to beat these issues here we tend to area unit
introducing this project – Secured request of fabric. The most objective of this project is to firmly communicate with the
assembly unit to the administration unit, sales unit, quality check unit and store house unit and to avoid information
escape. As information knowledge distributor has given sensitive data to purportedly trustworthy agents. Some
information has been leaked and located in Associate in nursing unauthorized place. The distributor should assess the
probability that the leaked information came from one or a lot of agents, as opposition having been gathered by
alternative freelance means that. Therefore we tend to style in such the way that the information on reaching any
destination says agent or unauthorized party, information in addition because the scientific discipline address of the
receiver can reach the distributor. If the distributor receives Associate in nursing scientific discipline address alternative
Associate in nursing agent’s address, the distributor finds out that the information has been leaked. Pretend object is
inserted beside the information at the time of distribution for police work the guilty agent. From the information
received, the distributor compares together with his information and finds out that agent has leaked {the information
the info the information} by scheming the chance for every agent on the leaked data.
Cloud computing, BOM, Security
R.Nallakumar, Dr. N. Sengottaiyan, L.S.Santhoshkumar
In various tasks of knowledge extraction such as relation extraction, community mining and the data extraction
semantic similarity between words is an important challenging task. Manually maintaining ontology will be difficult when the
domain changes over time. For each new word new meanings will be assigned to those words. So automatic methods to
estimate semantic similarity between words using web search interface is required. In existing systems Page counts and snippets
are the two information sources provided by the web search engines. But page counts are not sufficient for measuring semantic
similarity. The method done earlier is SWD(Semantic Word Distance) and it was calculated using page count and the SVM
classifier is used to classify the synonymous and non-synonymous words to find out the semantic similarity among words.
The web documents are retrieved from the query and the feature selection are extracted from the documents. Paper proposes
to use Hierarchical Agglomerative clustering(HAC) to group the documents in to different clusters. The query word will be
passed to the word net to extract the synonym for those words X and Y. The extracted synonym will be compared to the feature
selection of the web documents. Using Naive Bayes classifier the probability value will be calculated for the semantically related
word.
Knowledge extraction, semantic, agglomerative clustering