Debahuti Mishra

A Naïve SVM-KNN based stock market trend reversal analysis for Indian benchmark indices

This paper proposes a hybridized framework of Support Vector Machine (SVM) with K-Nearest Neighbor approach for Indian stock market indices prediction. The objective of this paper is to get in-depth knowledge in the stock market in Indian Scenario with the two indices such as, Bombay Stock Exchange (BSE Sensex) and CNX Nifty using technical analysis methods and tools such as predicting closing price, volatility and momentum of the stock market for the available data. This hybrid model uses SVM with different kernel functions to predict profit or loss, and the output of SVM helps to compute best nearest neighbor from the training set to predict future of stock value in the horizon of 1 day, 1 week and 1 month. The proposed SVM and KNN based prediction model is experienced with the above mentioned distinguished stock market indices and the performance of proposed model has been computed using Mean Squared Error and also been compared with recent developed models such as FLIT2NS and CEFLANN respectively. The limitation of both of those existing models undergoes complex weight updating procedures, whereas, proposed SVM-KNN hybridized model scales relatively well to high dimensional data and the trade-off between classifier complexity and error can be controlled explicitly and have better prediction capability.

An enhanced classifier fusion model for classifying biomedical data

Classification is a technique where we discover the hidden class level of the unknown data. As different classification methods produces different accuracy according to the class level; classifier fusion is the solution to achieve more accuracy in every level of the input data. Selection of a suitable classifier in classifier fusion is a tedious task. In the proposed model, the output of the three classifiers is fed to the dynamic classifier fusion technique. This model will use each classifier for every individual data. We have used principal component analysis (PCA) to deal with issues of high dimensionality in biomedical classification. Three types of classification techniques on microarray data like multi layer perceptron (MLP), FLANN and PSO-FLANN have been implemented and compared; it has been observed that MLP is showing better result. We have also proposed a model for classifier fusion, where the model will choose the relevant classifiers according to the different region of datasets.

Particle swarm optimization based fuzzy frequent pattern mining from gene expression data

The FP-growth algorithm is currently one of the fastest approaches to frequent item set mining. Fuzzy logic provides a mathematical framework where the entire range of the data lies in between 0 and 1. The PSO algorithm was developed from observations of the social behavior of animals, including bird flocking and fish schooling. It is easier to implement than evolutionary algorithms because it only involves a single operator for updating solutions. In contrast, evolutionary algorithms require a particular representation and specific methods for cross-over, mutation, and selection. Furthermore, PSO has been found to be very effective in a wide variety of applications, being able to produce good solutions at a very low computational cost. In this paper, we have considered the fuzzified dataset and have implemented various frequent pattern mining techniques. Out of the various frequent pattern mining techniques it was found that Frequent Pattern (FP) growth method yields us better results on a fuzzy dataset. Here, the frequent patterns obtained are considered as the set of initial population. For the selection criteria, we had considered the mean squared residue score rather using the threshold value. It was observed that out of the four fuzzy based frequent mining techniques, the PSO based fuzzy FP growth technique finds the best individual frequent patterns. Also, the run time of the algorithm and the number of frequent patterns generated is far better than the rest of the techniques used.

FPGA Implementation of Software Defined Radio-Based Flight Termination System

This paper proposes a field-programmable gate array (FPGA)-based software defined radio (SDR) implemented flight termination system (FTS). This is purely a new kind of implementation of digital FTS in SDR platform. The applied design procedure replaces a multiple platform-based system with a single platform. It also guarantees reconfigurable, interoperable, portable, and handy FTS, and maintains errorless, bug free, and reliable implementation. Real-time flight termination operation demands a very highly reliable and ruggedized platform. Hence, the FTS is implemented in FPGA. In order to minimize hardware resources and to enable future upgradation, efficient optimization technique has been applied. LabVIEW, a high-level programming language is used to simulate and implement the system in real time and enables rapid prototyping. The system was validated at subsystems level by measurements of different parameters in various intermediate stages of processing, and further was validated as an integrated system at real-time telecommunication operation environment.

Improved search technique using wildcards or truncation

Search engine technology plays an important role in web information retrieval. However, with Internet information explosion, traditional searching techniques cannot provide satisfactory result due to problems such as huge number of result Web pages, unintuitive ranking etc. Therefore, the reorganization and post-processing of Web search results have been extensively studied to help user effectively obtain useful information. This paper has basically three parts. First part is the review study on how the keyword is expanded through truncation or wildcards (which is a little known feature but one of the most powerful one) by using various symbols like * or! .The primary goal in designing this is to restrict ourselves by just mentioning the keyword using the truncation or wildcard symbols rather than expanding the keyword into sentential form. Second part consists of the review on subdivision based on wildcards. It is based on the observation that documents are often found to contain terms with high information content which summarize their subject matter. The third part consists of a proposed algorithm based on the above two. The main goal of this paper is to develop a very efficient search technique by which the information retrieval will be very fast, reducing the amount of extra labor needed on expanding the query.

Fuzzy pattern tree approach for mining frequent patterns from gene expression data

Frequent pattern mining has been a focused theme in data mining research for over a decade. A lot of literature has been dedicated to this research and huge amount of work has been made, ranging from efficient and scalable algorithms for frequent item set mining in transaction databases to numerous research frontiers. Frequent pattern mining (FPM) has been applied successfully in business and scientific data for discovering interesting association patterns, and is becoming a promising strategy in microarray gene expression analysis. As we know, Fuzzy logic provides a mathematical framework that is compatible with poorly quantitative yet qualitatively significant data. In this paper, we have fuzzified our original dataset and have applied various frequent pattern mining techniques to discover meaningful frequent patterns. Also, we have drawn a clear comparison of the frequent pattern mining techniques in the original and the fuzzified data in terms of parameters like runtime of the algorithm and the number of frequent patterns generated. As a result, it was found that the fuzzified set is capable of discovering a large number of frequent patterns and have a better running time capability.

A signal-to-noise classification model for identification of differentially expressed genes from gene expression data

A major focus in cancer research is identifying genetic markers or biomarkers. To build a robust classifier we have to find out the differentially expressed genes (key genes) in binary classification. The differentially expressed genes or biomarker gene selection is the preprocessing task for cancer classification. In this paper, we have compared the results of two approaches for selecting biomarkers from Leukemia data set. The first approach for feature selection is by implementing k-means clustering and signal-to-noise ratio (SNR) method for gene ranking, the top scored genes from each cluster is selected and given to the classifiers. The second approach uses signal to noise ratio ranking only for feature selection. For validation of both the approaches, we have used k nearest neighbor (kNN), support vector machine (SVM), probabilistic Neural Network (PNN) and Feed Forward Neural Network (fNN). After comparing the final results of two approaches we have got 100%, 96%and 96% accuracy with SVM, kNN and PNN respectively in first approach with five numbers of genes. Whereas, performance of FNN is 2.17 with 10 numbers of genes. In second approach we have got 96%, 96% and 62% accuracies for SVM, kNN and PNN respectively for 5 numbers of genes and the performance of FNN is 2.52 for 10 genes.

Gene expression network discovery: a pattern based biclustering approach

Discovering biologically significant information from gene expression data is now a days playing important role in gene function detection, gene regulation, drug discovery, detecting and predicting the diseases. Many traditional clustering algorithms are present to discover such gene regulations. Such discovered clusters are known as global clusters, which incurs more processing overhead. To overcome such problem, the biclustering approach, also known as local clustering has been emerged. Generally, there are two ways of measuring the similarity among subset of objects and attributes. First one is grouping the data elements by measuring the similarity based on the proximity. But, there may be the case that, many objects and attributes which are far apart but the gives significant meaning for being grouped. This problem can be solved by the second method, which not only measures the proximity of data elements but also find subset of objects and attributes which forms similar or coherent patterns such as scaling and shifting irrespective of their proximity. In this paper, we have implemented the pattern based clustering and before that the dimensionality reduction using Principal Component Analysis (PCA) is used to reduce the attributes without loss of information. We have compared the Minimum Squared Residue (MSR) approach of Cheng and Church with our proposed model. Our method shows its better performance as compared to MSR based approach.

A novel approach for selecting informative genes from gene expression data using Signal-to-Noise Ratio and t-statistics

Signal-to-Noise Ratio (SNR) and t-statistics are widely used for gene ranking in the analysis of microarray gene expression data. By implementing these filtering techniques directly to the microarray data may give redundant features, as we may have redundant expression values of number of genes in the data set. By grouping the genes bearing similar expression values in a single cluster and then implementing the given filtering techniques to rank the genes in each cluster and by selecting top ranked genes from each cluster give better result towards biomarker selection. In this paper we have taken four cancer data sets and k-means clustering technique to cluster the genes. Support vector machine and k-nearest Neighbor are used for classification and the method for validation is 10 fold cross validation.

Discovering non-exclusive functional modules from gene expression data

Biological processes are not independent of each other as genes participate in multiple different processes. Each gene should be assigned to multiple biclusters. In real life, more than one gene is responsible for a particular type of disease. The biclustering can associate clusters with gene arrangement patterns, preserving genomic information. Additionally, overlapping capability is desirable for the discovery of multiple conserved patterns within a single genome. In strict or crisp partition-based biclustering, each gene/condition belongs to exactly one functional module whereas, addressing some biological questions requires partitioning methods leading to non-exclusive functional modules. The proposed method involves a novel strategy to discover such non-exclusive pattern-based biclusters using fuzzy set approach. We have evaluated the performance of our proposed model with few existing ones and the result shows that this can be suitable for application to genomes with high genetic exchange and various conserved gene arrangements in gene regulatory networks.