About Me |
---|
To succeed in an environment of growth and excellence and earn a job which provides me job satisfaction and self development and help me achieve personal as well as organizational goals. |
Worked on Image Processing, Sentiment Analysis, Optimization and data classification .
Sl. No. | Title | Issuer |
---|
Backpropagation (BP) algorithms are widely used for propagating adaptive learning to an artificial neural network (ANN), which uses a gradient descent learning approach to train an ANN. But choosing an efficient and appropriate training methodology is still a challenge. The gravitational search algorithm (GSA) is a recently developed metaheuristic approach which tries to find out a global solution by balancing both exploration and exploitation. Though GSA always is guaranteed to provide a better solution, some time it takes more time to generate a solution. In contrary to this, gradient descent-based algorithms have the limitation of getting stuck into local optima. In view of this, the proposed work has tried to do an amalgamation of both the techniques to avoid the individual deficiencies lying in each of the techniques. A synergy of both the techniques is validated by some of the benchmark datasets along with using some benchmark functions. The performance is measured and compared by running the algorithms individually and by the hybrid approach. The generated results establish the validity of the proposed model.
Lots of research has been carried out globally to design a machine classifier which could predict it
from some physical and bio-medical parameters. In this work a hybrid machine learning classifier has
been proposed to design an artificial predictor to correctly classify diabetic and non-diabetic people.
The classifier is an amalgamation of the widely used K-means algorithm and Gravitational search
algorithm (GSA). GSA has been used as an optimization tool which will compute the best centroids
from the two classes of training data; the positive class (who are diabetic) and negative class (who
are non-diabetic). In K-means algorithm instead of using random samples as initial cluster head, the
optimized centroids from GSA are used as the cluster centers. The inherent problem associated with
k-means algorithm is the initial placement of cluster centers, which may cause convergence delay
thereby degrading the overall performance. This problem is tried to overcome by using a combined
GSA and K-means.
The paper contains an extensive experimental study which focuses on a major idea on Target
Optimization (TO) prior to the training process of artificial machines. Generally, during training
process of an artificial machine, output is computed from two important parameters i.e. input and
target. In general practice input is taken from the training data and target is randomly chosen, which
may not be relevant to the corresponding training data. Hence, the overall training of the neural
network becomes inefficient. The present study tries to put forward TO as an efficient methodology
which may be helpful in addressing the said problem. The proposed work tries to implement the
concept of TO and compares the outcomes with the conventional classifiers. In this regard, different
benchmark data sets are used to compare the effect of TO on data classification by using Particle
Swarm Optimization (PSO) and Gravitational Search Algorithm (GSA) optimization techniques.
In this paper, we propose a particle swarm-based extreme learning
machine (ELM) to classify datasets with varying number of classes. This work
emphasises on a couple of important parameters, like maximisation of
classification accuracy and minimisation of training time. As a machine
classifier, an ELM has been chosen, which is an improvement over back
propagation network. For each of the input dataset an optimised target was
determined by using particle swarm optimisation (PSO) technique. Those
specific targets are used with the input data to train the ELM during
classification process. For this, some of the bench mark classification datasets
are used. To compare the proposed method and some of the existing methods
an extensive experimental study has been carried out; a comparative analysis is
done by taking parameters like percentage of classification accuracy, training
time and complexity of the computing algorithm.
Classification is one of the most active research and application areas of artificial neural networks (ANN). One of the difficulties in using ANN is to find the most suitable combination of training, learning and transfer function for classification of data sets with increasing number of features and classified sets. In this paper we have studied the effect of different combinations of functions while using artificial neural network as a classifier and analyzed the suitability of these functions for different kinds of datasets. The appropriateness of the proposed work has been determined on the basis of mean square error, rate of convergence, and accuracy of the classified dataset. Our inferences are based on the simulation results over the datasets used.
Data Classification and predictions are one of the prime tasks in Data mining. They continue to play a vital role in the area of computer science and data processing field. Clustering and classifications in Data Mining are used in various domains to give meaning to the available data and give some useful prediction results which can be applied to some of the crucial problem areas of the real world. Diabetes mellitus otherwise known as a slow poison by the medical experts is a major, alarming and gradually becoming a global problem. This paper experimented and used the concept of modified extreme learning machine to identify the patients of being diabetic or non-diabetic basing on some previously given data which in turn helps the medical people to identify whether someone is affected by diabetes or not.
In this paper we propose a particle swarm based back propagation neural network model which uses an optimized target to maximize the classification accuracy of the classifier. By using Particle swarm optimization technique an optimized target for each class was determined and there after the artificial neural network is used to classify the data using these targets. For this, some of the bench mark classification datasets are used, which are taken from UCI learning repository. An extensive experimental study has been carried out to compare the proposed method and existing method on the same datasets and a comparative analysis is done by taking several parameters like percentage of accuracy, time of response and complexity of the algorithm. During this study we have examined the performance improvement of the proposed PSO and BPN combined approach over the conventional BPN approach to generate classification inferences from the training and testing results.