Qualification: 
M.Tech, B-Tech
prathibhamolcp@am.amrita.edu

Prathibhamol C. P. currently serves as an Assistant Professor (Senior Grade) at the Department of Information Technology at Amrita School of Engineering, Amritapuri.

Publications

Publication Type: Conference Paper

Year of Publication Title

2017

Prathibhamol CP, Suresh, A., and Suresh, G., “Prediction of cardiac arrhythmia type using clustering and regression approach (P-CA-CRA)”, in 2017 International Conference on Advances in Computing, Communications and Informatics (ICACCI), Udupi, India, 2017.[Abstract]


Cardiac Arrhythmia is a disease dealing with improper beating of heart. The improper condition may be fast beating or slow beating associated with heart. This paper proposes a detection or prediction scheme in the type of cardiac arrhythmia disease. It uses a clustering approach and regression methodology. The clustering approach used is DBSCAN and for regression, multiclass logistic regression is employed. By performing DBSCAN clustering algorithm, the whole dataset is disintegrated into disjoint clusters. Those clusters which are found to contain less instances, are then taken for consideration. These clusters are subjected to multiclass logistic regression. This is because, clustering approach is an unsupervised process. Once regression is performed, we have reached at a conclusion, about what type of cardiac arrhythmia it is. The proposed method achieves an overall accuracy of 80%, when compared with various other existing approaches.

More »»

2017

S. Athira, Poojitha, K., and Prathibhamol CP, “An efficient solution for multi-label classification problem using apriori algorithm (MLC-A)”, in 2017 International Conference on Advances in Computing, Communications and Informatics (ICACCI), Udupi, India, 2017.[Abstract]


Recently, multi-label classification has gained prime importance among the classification problems. The applications of classification problems has increased so rapidly that the need for efficient and accurate classifiers has become a vital requirement in the area of data mining. Multi-label classification problem is distinguished from the single label classification because of the capability to handle multiple labels. In this paper, we put forward a good method to predict the multiple labels of an unlabeled instance by using apriori algorithm. The proposed solution MLC-A is able to find unknown labels for any test instance by checking for the presence of particular set of attributes, along with their discretized values. MLC-A intends to employ Apriori algorithm for multi-label classification purpose, in contrast, to the well known existing methods. Apriori algorithm is utilized to find the relationship between attributes and labels by generating rules. The generated rules will find the occurrence of a label given the existence of various attributes. The experimental results on datasets of yeast, scene and emotions had given excellent accuracy. Using MLC-A the time complexity has reduced to a great extend. The accuracy and efficiency of the proposed method is evaluated by using hamming loss and is compared to the state of art.

More »»

2016

Prathibhamol CP, Jyothy, K. V., and Noora, B., “Multi label classification based on logistic regression (MLC-LR)”, in Second International Symposium on Emerging Topics in Computing and Communications, International Conference on Advances in Computing, Communications and Informatics (ICACCI), Jaipur, India, 2016.[Abstract]


Numerous class labels associated with each data instance is the main feature of any multi-label classification (MLC) problem. Correct prediction of class labels related to any test data is a big challenge in this domain. MLC can be applied in many fields such as personality prediction, cancer prognosis prediction, image annotation etc. In this paper(MLC-LR), we have employed problem transformation method for solving MLC. The proposed method uses initially clustering in the feature space. It is then followed by FP-growth algorithm for finding the relationship between labels. Once the desired clusters are obtained, then normalization of data associated with each cluster is performed. Also logistic regression is then applied over the normalized data for each particular cluster pertaining to all labels. When a new instance arrives in the testing phase, immediately the nearby cluster is identified by means of Euclidean distance metric as the measure. Rules related to label space for the nearby cluster is extracted to check for hypothesis of each antecedent label. If the calculated value is higher than a predefined threshold, it is assumed that both antecedent and consequent labels as the estimated labels for that test instance.

More »»

2016

Prathibhamol CP, Amala, G. S., and Kapadia, M., “Anomaly detection based multi label classification using Association Rule Mining (ADMLCAR)”, in Second International Symposium on Emerging topics in Computing Communication, International Conference on Advances in Computing, Communications and Informatics (ICACCI), Jaipur, India, 2016.[Abstract]


Multi label classification contains multiple labels in the label space. Any Multi label classification problem (MLC) deals with numerous class labels associated with data instances. Due to this, correct prediction of labels for a test data remains as a challenge in this field. In this paper an Anomaly Detection based Multi Label Classification using Association Rule Mining (ADMLCAR) is used for solving MLC problem. Conventionally, most of the multi label classification problem is solved by either of the two methods: Problem transformation, Algorithm adaptation. But the method discussed in this paper aims at a novel method different from traditional solution to multi label classification problem. For clustering, ADMLCAR uses k-means algorithm and for association rule mining purpose it uses vertical data format. To predict the test data instances, this method locates the nearest cluster. Once the clusters are identified it uses oversampling principal component analysis (PCA) within the nearest cluster with respect to test instances. Oversampling PCA is used to emphasize the need for confirming the fact that test instance's label set will not only be confined to its nearest cluster label set. This is because, anyways the test instance will be identified to a nearest cluster by means of Euclidean distance measure but as clustering is unsupervised the nearest cluster may contain many objects entities of different label sets.

More »»

2016

H. Haripriya, Prathibhamol CP, Pai, Y. R., Sandeep, M. S., Sankar, A. M., a, S. N. V., and Prof. Prema Nedungadi, “Multi label prediction using association rule generation and simple k-means”, in 2016 International Conference on Computational Techniques in Information and Communication Technologies (ICCTICT), 2016.[Abstract]


Lately, modern applications like information retrieval, semantic scene classification, music categorization and functional genomics classification highly require multi label classification. A rule mining algorithm apriori is widely used for rule generation. But Apriori is used many times on categorical data, it is seldom used for numerical data. This leads to an idea that with proper data pre-processing, a lot of intangible rules can be derived from such numerical datasets. Since the algorithm will check each and every datasets, we used a simple k-means clustering approach for dividing the processing space of Apriori and thus rules are generated for each cluster. The accuracy of the algorithm is calculated using hamming loss and is presented in the paper. This hybrid algorithm directly aims to find out hidden patterns in huge numerical datasets and make reliable label prediction easier.

More »»
PDF iconmulti-label-prediction-using-association-rule-generation-and-simple-k-means.pdf

Publication Type: Conference Proceedings

Year of Publication Title

2016

Prathibhamol CP and Asha Ashok, “Solving multi label problems with clustering and nearest neighbor by consideration of labels”, 2nd International Symposium on Signal Processing and Intelligent Recognition Systems (SIRS'15), Advances in Intelligent Systems and Computing, vol. 425. Springer , pp. 511-520, 2016.[Abstract]


In any Multi label classification problem, each instance is associated with multiple class labels. In this paper, we aim to predict the class labels of the test data accurately, using an improved multi label classification approach. This method is based on a framework that comprises an initial clustering phase followed by rule extraction using FP-Growth algorithm in label space. To predict the label of a new test data instance, this technique searches for the nearest cluster, thereby locating k-Nearest Neighbors within the corresponding cluster. The labels for the test instance are estimated by prior probabilities of the already predicted labels. Hence, by doing so, this scheme utilizes the advantages of the hybrid approach of both clustering and association rule mining.The proposed algorithm was tested on standard multi label datasets like yeast and scene. It achieved an overall accuracy of 81% when compared with scene dataset and a 68% in yeast dataset. © Springer International Publishing Switzerland 2016.

More »»
Faculty Research Interest: