International Journals

Permanent URI for this collectionhttps://dspace.psgrkcw.com/handle/123456789/164

Browse

Now showing 1 - 20 of 68

MACHINE LEARNING APPROACH FOR PREOPERATIVE ANAESTHETIC RISK PREDICTION
(Academy Publishers, Finland, 2009-05) Karpagavalli S; Jamuna K S; Vijaya M S
Risk is ubiquitous in medicine but anaesthesia is an unusual speciality as it routinely involves deliberately placing the patient in a situation that is intrinsically full of risk. Patient safety depends on management of those risks; consequently, anaesthetist has been at the forefront of clinical risk management. Anaesthetic risk classification is of prime importance not only in carrying out the day-to-day anaesthetic practice but coincides with surgical risks and morbidity condition. The preoperative assessment is made to identify the patients risk level based on American Society of Anesthesiologists (ASA) score that is widely used in anaesthetic practice. This helps the anaesthetist to make timely clinical decision. Machine Learning techniques can help the integration of computer-based systems in the healthcare environment providing opportunities to facilitate and enhance the work of medical experts and ultimately to improve the efficiency and quality of medical care. This paper presents the implementation of three supervised learning algorithms, C4.5 Decision tree classifier, Naive Bayes and Multilayer Perceptron in WEKA environment, on the preoperative assessment dataset. The classification models were trained using the data collected from 362 patients. The trained models were then used for predicting the anaesthetic risk of the patients. The prediction accuracy of the classifiers was evaluated using 10-fold cross validation and the results were compared.
A NOVEL APPROACH FOR PASSWORD STRENGTH ANALYSIS THROUGH SUPPORT VECTOR MACHINE
(Academy Publishers, Finland, 2009-11) Karpagavalli S; Jamuna K S; Vijaya M S
Passwords are ubiquitous authentication methods and they represent the identity of an individual for a system. Users are consistently told that a strong password is essential these days to protect private data. Despite the existence of more secure methods of authenticating users, including smart cards and biometrics, password authentication continues to be the most common means in use. Thus it is important for organizations to recognize the vulnerabilities to which passwords are subjected, and develop strong policies governing the creation and use of passwords to ensure that those vulnerabilities are not exploited. This work employs machine Learning technique to analyze the strength of the password to facilitate organizations launch a multi-faceted defense against password breach and provide a highly secure environment. A supervised learning algorithm namely Support Vector Machine is used for classification of password. The linear and nonlinear SVM classification models are trained using the features extracted from the password dataset. The trained model shows the prediction accuracy of about 98% for 10-fold cross validation
PROACTIVE PASSWORD STRENGTH ANALYZER USING FILTERS AND MACHINE LEARNING TECHNIQUES
(International Journal of Computer Applications, 2010-10) Suganya G; Karpagavalli S
Passwords are ubiquitous authentication methods and they represent the identity of an individual for a system. Users are consistently told that a strong password is essential these days to protect private data. Despite the existence of more secure methods of authenticating users, including smart cards and biometrics, password authentication continues to be the most common means in use. Thus it is important for organizations to recognize the vulnerabilities to which passwords are subjected, and develop strong policies governing the creation and use of passwords to ensure that those vulnerabilities are not exploited. This work proposes a framework to analyze the strength of the password proactively. To analyze the chosen password, filters and support vector machine are employed. This framework can be implemented as a submodule of the access control scheme.
A STUDY ON EMAIL SPAM FILTERING TECHNIQUES
(International Journal of Computer Applications, 2010-12) Christina V; Karpagavalli S; Suganya G
Electronic mail is used daily by millions of people to communicate around the globe and is a mission-critical application for many businesses. Over the last decade, unsolicited bulk email has become a major problem for email users. An overwhelming amount of spam is flowing into users’ mailboxes daily. Not only is spam frustrating for most email users, it strains the IT infrastructure of organizations and costs businesses billions of dollars in lost productivity. The necessity of effective spam filters increases. In this paper, we presented our study on various problems associated with spam and spam filtering methods, techniques.
AUTOMATIC SPEECH RECOGNITION: ARCHITECTURE, METHODOLOGIES, CHALLENGES - A REVIEW
(International Journal of Advanced Research in Computer Science, 2011-11) Karpagavalli S; Deepika R; Kokila P; Usha Rani K; Chandra E
For more than three decades, a great amount of research was carried out on various aspects of speech signal processing and its applications. Highly successful application of speech processing is Automatic Speech Recognition (ASR). Early attempts to ASR consisted of making deterministic models of whole words in a small vocabulary and recognizing a given speech utterance as the word whose model comes closest to it. The introduction of Hidden Morkov Models (HMMs) in the early 1980 provided much more powerful tool for speech recognition. And the recognition can be done for continuous speech using large vocabulary, in a speaker independent manner. Today many products have been developed that successfully utilize ASR for communication between human and machines. Performance of speech recognition applications deteriorates in the presence of reverberation and even low levels of ambient noise. Robustness to noise, reverberation and characteristics of the transducer is still an unsolved problem that makes the research in the area of speech recognition still very active. A detailed study on ASR carried out and presented in this paper that covers the basic model of speech recognition, applications
EMAIL SPAM FILTERING USING SUPERVISED MACHINE LEARNING TECHNIQUES
(International Journal of Advanced Research in Computer Science, 2011-12) Christina V; Karpagavalli S; Suganya G
E-mail spam, known as unsolicited bulk Email (UBE), junk mail, or unsolicited commercial email (UCE), is the practice of sending unwanted e-mail messages, frequently with commercial content, in large quantities to an indiscriminate set of recipients. Spam is prevalent on the Internet because the transaction cost of electronic communications is radically less than any alternate form of communication. There are many spam filters using different approaches to identify the incoming message as spam, ranging from white list / black list, Bayesian analysis, keyword matching, mail header analysis, postage, legislation, and content scanning etc. Even though we are still flooded with spam emails everyday. This is not because the filters are not powerful enough, it is due to the swift adoption of new techniques by the spammers and the inflexibility of spam filters to adapt the changes. In our work, we employed supervised machine learning techniques to filter the email spam messages. Widely used supervised machine learning techniques namely C 4.5 Decision tree classifier, Multilayer Perceptron, Naïve Bayes Classifier are used for learning the features of spam emails and the model is built by training with known spam emails and legitimate emails. The results of the models are discussed.
EMPIRICAL EVALUATION OF FEATURE SELECTION TECHNIQUE IN EDUCATIONAL DATA MINING
(ARPN Journal of Science and Technology, 2012) A S, Kavitha; J, VijiGrpisy; R, Kavitha
In machine learning the classification task is commonly referred to as supervised learning. In supervised learning there is a specified set of classes and objects are labeled with the appropriate class. The goal is to generalize from the training objects that will enable novel objects to be identified as belonging to one of the classes. Evaluating the performance of learning algorithms is a fundamental aspect of machine learning. The primary objective of this thesis is to study the classification accuracy using feature selection with machine learning algorithms. Feature selection is considered successful if the dimensionality of the data is reduced and accuracy of a learning algorithm improves or remains the same. Hence our contribution in this research is to prepare an educational dataset with real time feedback from students and try to apply the same with weka tool to measure the classification accuracy. Some part of implementation is compiled with weka, which is written in java and experiment with weka explorer.
A SURVEY ON SPECIES RECOGNITION SYSTEM FOR PLANT CLASSIFICATION
(International Journal of Computer Technology & Applications (IJCTA), 2012-05) Arunpriya C; Antony Selvadoss Thanamani
Several attempts have been made by taxonomists and morphometricians to find out the best automated identification of biological species, but they haven’t found any effective species recognition system for several decades. It would be very helpful to carry out the behavioral and ecological studies on plants for plant classification with the help of species recognition system. Each species of a plant and its leaf has its own distinctive patterns, which enabled the researchers to perform some research on it to accurately classify the plants. In general, plant species recognition system includes image categorization and object recognition. Plant species recognition system is completely different from common image categorization, since the variation between the one plant species leaf and other is very small. As a result, the traditional image categorization techniques do not perform effectively on plant images. Automatic plant recognition system has not yet been well established largely being the fact that lack of research in this field and the complexity in obtaining the database. Species recognition system on plants is one of the major concerns at present and there is huge requirement for several researches to deal with the better plant species recognition system. In this survey, several plant species recognition system are discussed which will show the way for development of better plant species recognition system for plant classification.
ISOLATED TAMIL DIGIT SPEECH RECOGNITION USING TEMPLATE-BASED AND HMM-BASED APPROACHES
(Springer, 2012-07) Karpagavalli S; Deepika R; Kokila P; Usha Rani K; Chandra E
For more than three decades, a great amount of research was carried out on various aspects of speech signal processing and its applications. Highly successful application of speech processing is Automatic Speech Recognition (ASR). Early attempts to ASR consisted of making deterministic models of whole words in a small vocabulary and recognizing a given speech utterance as the word whose model comes closest to it. The introduction of Hidden Markov Models (HMMs) in the early 1980 provided much more powerful tool for speech recognition. And the recognition can be done for continuous speech using large vocabulary, in a speaker independent manner. Two approaches like conventional template-based and Hidden Markov Model usually performs speaker independent isolated word recognition. In this work, speaker independent isolated Tamil digit speech recognizers are designed by employing template based and HMM based approaches. The results of the approaches are compared and observed that HMM based model performs well and the word error rate is greatly reduced.
SCALY NEURAL NETWORKS FOR SPEECH RECOGNITION USING DTW AND TIME ALIGNMENT ALGORITHMS
(International Journal of Scientific and Research Publications,, 2012-10) Sabitha P V; Karpagavalli S
Speech recognition has been an active research topic for more than 50 years. Interacting with the computer through speech is one of the active scientific research fields particularly for the disable community who face variety of difficulties to use the computer. Such research in Automatic Speech Recognition (ASR) is investigated for different languages because each language has its specific features. Neural Networks are, in essence, biologically inspired networks since they are based on the current understanding of the biological nervous system. In essence they are comprised of a network of densely interconnected simple processing elements, which perform in a manner analogous to the most development of neural networks, and a basic introduction to their theory is outlined in this elementary functions of a biological neuron. Reduced connectivity neural networks are discussed and the scaly architecture neural network is described. Various algorithms are available to perform this time alignment of the input pattern to the neural network and the performance of the neural network is dependent upon the performance of the time alignment algorithm used. In this chapter, the various types of time alignment algorithms are described and their operation is outlined in detail.
ISOLATED TAMIL WORDS SPEECH RECOGNITION USING LINEAR PREDICTIVE CODING AND NEURAL NETWORKS
(2012-12) Sabitha P V; Karpagavalli S
Speech Recognition is the ability of a computer to recognize general, naturally flowing utterances from a wide variety of users. In recent years, with the new generation of computing technology, speech technology becomes the next major innovation in man-machine interaction. Automatic Speech Recognition (ASR) system takes a human speech utterance as an input and returns a string of words as output. Research on speech recognition has led to variety of applications like hands free and eyes free applications, voice user interfaces, simple data entry, forensic applications, voice authentication, biometrics, robotics, air traffic controllers, preparation of medical reports, learning tools for handicapped, and reading tools for blind people. Even though research in speech recognition in English language attained certain maturity, speech interfaces in Indian languages still in the startup level. Tamil is one of the widely spoken Indian languages of the world with more than 77 million speakers. Speech interfaces in Indian languages will enable the people in various semiurban and rural parts of India to use telephones and Internet services. In the proposed work, isolated Tamil words speech recognition interface is designed using neural network algorithm. To design the system, a dataset of 10 Tamil words uttered by 20 speakers each word 5 times has been prepared. Linear predictive coding of order 8 is used for feature extraction. Back-propagation training is carried with the feature vectors extracted using LPC from the speech files in the dataset. Multilayer Perceptron algorithm in neural network is employed for recognition of the words using the trained model. An interface also designed to recognize the Tamil words uttered by the user. The average recognition rate of the system is 93.6% and for few words it gives 100% accuracy. The performance of the system is measured using word recognition rate and word error rate
EMPIRICAL EVALUATION OF FEATURE SELECTION TECHNIQUE IN EDUCATIONAL DATA MINING
(ARPN Journal of Science and Technology, 2012-12) A S, Kavitha; R, Kavitha; J, Viji Gripsy
In machine learning the classification task is commonly referred to as supervised learning. In supervised learning there is a specified set of classes and objects are labeled with the appropriate class. The goal is to generalize from the training objects that will enable novel objects to be identified as belonging to one of the classes. Evaluating the performance of learning algorithms is a fundamental aspect of machine learning. The primary objective of this thesis is to study the classification accuracy using feature selection with machine learning algorithms. Feature selection is considered successful if the dimensionality of the data is reduced and accuracy of a learning algorithm improves or remains the same. Hence our contribution in this research is to prepare an educational dataset with real time feedback from students and try to apply the same with weka tool to measure the classification accuracy. Some part of implementation is compiled with weka, which is written in java and experiment with weka explorer.
DISCOVERING TAMIL WRITER IDENTITY USING GLOBAL AND LOCAL FEATURES OF OFFLINE HANDWRITTEN TEXT
(International Review on Computers and Software (IRECOS), 2013) Thendral T; Vijaya M S; Karpagavalli S
Writer identification is the process of identifying the individual based on their handwriting. Handwriting exhibits behavioral characteristics of an individual and has been considered as unique. The style and shape of the letters written vary slightly for same writer and entirely different for different writers. Also alphabets in the handwritten text may have loops, crossings, junctions, different directions etc. Hence accurate prediction of individual based on his/her handwriting is highly complex and challenging task. This paper proposes a new model for discovering the writer’s identity based on Tamil handwriting. Writer identification problem is formulated as classification task and a pattern classification technique namely Support Vector Machine has been employed to construct the model. It has been reported about 93.8% of prediction accuracy by RBF kernel based classification model.
SURVEY ON FUTURISTIC ALIGNMENT OF IT PROCESSES REFERENCE TO INDIA
(International Journal of Emerging Trends & Technology in Computer Science (IJETTCS), 2013-01) A S, Kavitha; J, Viji Gripsy; N, Deepa
Today environmental sustainability is a global issue. Main attention of Environmental sustainability is to preserve the capability of environment to support human life. It not only involves reduction of waste but also make waste as useful resource. In India green IT sustainability initiatives will double within 5 years. In 2010 it was $35 billion but in 2015 it will increase to $70 billion. Recently Indian IT companies such as HP, Dell, Acer, Wipro, HCL, Infosys and a few others are adopting green computing. And they are also encouraging their customers and employees to adopt green computing. We mainly focus on the minimization of power consumption and heat reduction.
AN EFFICIENT LEAF RECOGNITION ALGORITHM FOR PLANT CLASSIFICATION USING KERNELIZED SUPPORT VECTOR MACHINE
(International Journal of Computer Science and Management Research (IJCSMR), 2013-02) Arunpriya C; Antony Selvadoss Thanamani
Plant recognition has become an active area of research as most of the plant species are at the risk of extinction. This paper uses an efficient machine learning approach for the classification purpose. This proposed approach consists of three phases such as preprocessing, feature extraction and classification. The preprocessing phase involves a typical image processing steps such as transforming to gray scale and boundary enhancement. The feature extraction phase derives the common DMF from five fundamental features. The values of the optimized σ are then used as a gauge for variable selection. In this study Kernelized (K-SVM) model is applied to several benchmark data sets in order to estimate the effectiveness of the second-order sigma tuning procedure for an RBF kernel.12 leaf features which are extracted and orthogonalized into 5 principal variables are given as input vector to the K-SVM. Classifier tested with flavia dataset and a real dataset and compared with k-NN approach, the proposed approach produces very high accuracy and takes very less execution time.
FUZZY SHRINK THRESHOLDING BASED TEA LEAF IMAGE ENHANCEMENT USING WAVELET TRANSFORM
(International Journal of Computer Applications (IJCA), 2013-02) Arunpriya C; Antony Selvadoss Thanamani
In this paper a wavelet shrinkage algorithm based on fuzzy logic is proposed to improve the tea leaf image. The Tea Leaf images are normally changes to unclear images by the presence of noise, low or high dissimilarity both in the edge area and also in the image area. The Fuzzy shrink is used to enhance the image. In exacting, intra-scale dependency within wavelet coefficients is modeled using a fuzzy characteristic. This characteristic space distinguishes between significant coefficients, which depends on image discontinuity and noisy coefficients. This fuzzy characteristic is used for enhancing wavelet coefficients' information in the shrinkage step in this paper. Then a fuzzy membership function known as the spline-based curve is used to shrinks the wavelet coefficients based on the fuzzy characteristic. Here by using the interrelation between different channels as a fuzzy characteristic for improving the denoising performance compared to denoising each channel, separately. Examine the image denoising algorithm in the dual-tree discrete wavelet transform, which is the latest shiftable and customized version of discrete wavelet transform. Extensive comparisons with the high-tech image denoising algorithm indicate that the image denoising algorithm has a better performance in noise suppression and edge preservation as compared with the other methods. The spline based curve of a fuzzy membership function is more efficient one
A SURVEY ON PALM PRINTS AUTHENTICATION SYSTEM
(International Journal of Scientific Research in Computer Science Applications and Management Studies, 2013-07) Krishnaveni M; Arunpriya C
Biometrics is rapidly evolving technology which is being widely used in forensics such as criminal identification and top-security prison, and has the potential to be used in a large range of civilian application area. Reliability in computer aided individual authentication is becoming increasingly important in the information-based world, for effective protection system. Biometrics is physiological character of human beings, unique for every person that are frequently time invariant and easy to obtain. Palm print is one of the relatively new physiological biometrics due to its stable and unique characteristics. The quality information of palm print offers one of the powerful means in personal identification.
BREAST CANCER CLASSIFICATION USING SUPPORT VECTOR MACHINE AND GENETIC PROGRAMMING
(International Journal of Innovative Research in Computer and Communication Engineering (IJIRCCE), 2013-09) Menaka K; Karpagavalli S
Breast cancer is one of the most leading causes of death among women. The early detection of abnormalities in breast enables the radiologist in diagnosing the breast cancer easily. Efficient tools in diagnosing the cancerous breast will help the medical experts in accurate diagnosis and timely treatment to the patients. In this work, experiments carried out using Wisconsin Diagnosis Breast Cancer database to classify the breast cancer either benign or malignant. Supervised learning algorithm Support Vector Machine (SVM) with kernels like Linear, Polynomial and Radial Basis Function and evolutionary algorithm Genetic Programming are used to train the models. The performance of the models are analysed where genetic programming approach provides more accuracy compared to Support Vector Machine in the classification of breast cancer and seems to be an fast and efficient method.
ENHANCED SENTENCE-LEVEL TEXT CLUSTERING USING SEMANTIC SENTENCE SIMILARITY FROM DIFFERENT ASPECTS
(International Journal of Computer Science and Information Technologies, 2014) Saranya J; Arunpriya C
Sentence clustering plays a significant role in many text processing activities. For instance, several authors have discussed that integrate sentence clustering into extractive multi document summarization useful to address issues of content overlap, leading to better coverage. Existing work proposed fuzzy clustering algorithm which is used for relational input data. This existing algorithm uses a graph representation of the data, and performs based on Expectation-Maximization framework. Proposed system improves the result of the clustering by introducing the novel sentence similarity technique. In our proposed system we are propose a new way to determine sentence similarities from different aspects. Probably based on information people can obtain from a sentence, which is objects the sentence describes, properties of these objects and behaviors of these objects. Four aspects, Objects-Specified Similarity, Objects-Property Similarity, Objects-Behavior Similarity and Overall Similarity are calculated to estimate the sentence similarities. First, for each sentence, all nouns in noun phrases are chosen as the objects specified in the sentence, all adjectives and adverbs in noun phrases as the objects properties and all verb phrases as the objects behaviors. Then, the four similarities are calculated based on a semantic vector method. We also conducted an experimental study with that could help us to efficiently clustering the sentence level text. Our study shows that this algorithm generates better quality clusters than traditional algorithms; in other words, it is benefits to increase the accuracy of the clustering result.
AN EFFECTIVE TEA LEAF RECOGNITION ALGORITHM FOR PLANT CLASSIFICATION USING IMPROVED ANFIS ALGORITHM
(European Journal of Scientific Research, 2014) Arunpriya C; Antony Selvadoss Thanamani
A leaf is an organ of a vascular plant, as identified in botanical terms, and in particular in plant morphology. Naturally a leaf is a thin, flattened organ bear above the ground and it is mainly used for photosynthesis. Recognition of plants has become an active area of research as most of the plant species are at the risk of extinction. Most of the leaves cannot be recognized easily since some are not flat (e.g. succulent leaves and conifers), some does not grow above ground (e.g. bulb scales), and some does not undergo photosynthetic function (e.g. cataphylls, spines, and cotyledons). In this paper we have attempted to identify tea plant cultivars using classification techniques. Tea leaf images are loaded from digital cameras or scanners in the system. This proposed approach consists of three phases such as preprocessing, feature extraction and classification to process the loaded image. The tea leaf images can be identified accurately in the preprocessing phase by fuzzy denoising using Dual Tree Discrete Wavelet Transform (DT-DWT) in order to remove the noisy features and boundary enhancement to obtain the shape of leaf accurately. In the feature extraction phase, Digital Morphological Features (DMFs) are derived to improve the classification accuracy. Improved Adaptive Neuro-Fuzzy Inference System (ANFIS) is used for efficient classification. The ANFIS is trained by 60 tea leaves to classify them into 6 types. Experimental results proved that the proposed method classifies the tea leaves with more accuracy in less time. Thus, the proposed method achieves more accuracy in identifying the leaf type.