Parul Agarwal

Bio: Parul Agarwal is an academic researcher from Jaypee Institute of Information Technology. The author has contributed to research in topics: Clustering high-dimensional data & Canopy clustering algorithm. The author has an hindex of 6, co-authored 15 publications receiving 129 citations.

Nature-Inspired Algorithms: State-of-Art, Problems and Prospects

Abstract: Nature-inspired algorithms have gained immense popularity in recent years to tackle hard real world (NP hard and NP complete) problems and solve complex optimization functions whose actual solution doesn’t exist. The paper presents a comprehensive review of 12 nature inspired algorithms. This study provides the researchers with a single platform to analyze the conventional and contemporary nature inspired algorithms in terms of required input parameters, their key evolutionary strategies and application areas. A list of automated toolboxes available for directly evaluating these nature inspired algorithms over numerical optimization problems indicates the need for unified toolbox for all nature inspired algorithms. It also elucidates the users with the minimum and maximum dimensions over which these algorithms have already been evaluated on benchmark test functions. Hence this study would aid the research community to know what all algorithms could be examined for large scale global optimization to overcome the problem of ‘curse of dimensionality’.

Empirical analysis of five nature-inspired algorithms on real parameter optimization problems

Abstract: In the past few years nature-inspired algorithms are seen as potential tools to solve computationally hard problems. Tremendous success of these algorithms in providing near optimal solutions has inspired the researchers to develop new algorithms. However, very limited efforts have been made to identify the best algorithms for diverse classes of problems. This work attempts to assess the efficacy of five contemporary nature-inspired algorithms i.e. bat algorithm (BA), artificial bee colony algorithm (ABC), cuckoo search algorithm (CS), firefly algorithm (FA) and flower pollination algorithm (FPA). The work evaluates the performance of these algorithms on CEC2014 30 benchmark functions which include unimodal, multimodal, hybrid and composite problems over 10, 30, 50 and 100 dimensions. Control parameters of all algorithms are self-adapted so as to obtain best results over benchmark functions. The algorithms have been evaluated along three perspectives (a) statistical significance using Wilcoxon rank sum test (b) computational time complexity (c) convergence rate of algorithms. Experimental results and analysis revealed that ABC algorithm perform best for majority of the problems on high dimension, while on small dimension, CS is the best choice. FPA attain the next best position follow by BA and FA for all kinds of functions. Self adaptation of above algorithms also revealed the best values of input parameters for various algorithms. This study may aid experts and scientists of computational intelligence to solve intricate optimization problems.

Enhanced flower pollination algorithm on data clustering

Abstract: Nature-inspired algorithms are emerging as most compatible algorithms in obtaining near-optimal solution to complex problems. The ability of meta-heuristic algorithm to obtain global optimization solution largely depends on convergence behavior. In order to enhance convergence capability of latest nature-inspired algorithm i.e. flower pollination algorithm (FPA), a modified version is presented. The performance of modified FPA is tested over clustering application. Algorithm is assessed in contrast to bat algorithm, firefly algorithm, and conventional FPAs on 10 clustering data-sets. Out of 10 data-sets, 8 are derived from pattern recognition and 2 are artificially generated. Clustering results are computed in terms of objective function value and CPU time taken at each run. Run length distribution graphs illustrate the convergence behavior of algorithms. Results indicate that the proposed modified FPA outperforms its counterparts both in terms of attaining best fitness value and reducing the CPU ...

AD3S: Advanced Driver Drowsiness Detection System using Machine Learning

Abstract: Drivers drowsiness is one of the prime reasons for road accidents around the globe. Persistent monotonous driving for an extended period of time without rest leads to drowsiness and fatal mishaps. Automatic detection of driver's drowsiness can prevent a large number of road accidents and thus, can save valuable lives. In this work, an advanced system namely AD3S (Advanced Driver Drowsiness Detection System) using Android application has been developed. The system is capable of capturing real-time facial landmarks of the drivers. The facial landmarks are further utilized to compute several parameters namely Eye Aspect Ratio (EAR), Nose Length Ratio (NLR) and Mouth Opening Ratio (MOR) based on adaptive threshold which are capable of detecting driver's drowsiness. The highlighting feature of AD3S is that it is non-intrusive in nature and is cost effective. To test the efficacy of AD3S, machine learning and deep learning techniques have been applied over a data set of 1200 application users. The empirical results demonstrate that the proposed system is capable of detecting driver's drowsiness with an accuracy of approximately 98% with Bagging classifier.

[IEEE IEEE International Conference on Networking, Sensing and Control, 2004 - Taipei, Taiwan (March 21-23, 2004)] IEEE International Conference on Networking, Sensing and Control, 2004 - Particle swarm optimization algorithm and its application to clustering analysis

Abstract: Clustering analysis is applied generally to Pattern Recognition, Color Quantization and Image Classification. It can help the user to distinguish the structure of data and simplify the complexity of data from mass information. The user can understand the implied information behind extracting these data. In real case, the distribution of information can be any size and shape. A particle swarm optimization algorithm-based technique, called PSO-clustering, is proposed in this article. We adopt the particle swarm optimization to search the cluster center in the arbitrary data set automatically. PSO can search the best solution from the probability option of the Social-only model and Cognition-only model[1, 2, 3J. This method is quite simple and valid and it can avoid the minimum local value. Finally, the effectiveness of the PSO-clustering is demonstrated on four artificial data sets.

Flower pollination algorithm: a comprehensive review

Abstract: Flower pollination algorithm (FPA) is a computational intelligence metaheuristic that takes its metaphor from flowers proliferation role in plants. This paper provides a comprehensive review of all issues related to FPA: biological inspiration, fundamentals, previous studies and comparisons, implementation, variants, hybrids, and applications. Besides, it makes a comparison between FPA and six different metaheuristics such as genetic algorithm, cuckoo search, grasshopper optimization algorithm, and others on solving a constrained engineering optimization problem . The experimental results are statistically analyzed with non-parametric Friedman test which indicates that FPA is superior more than other competitors in solving the given problem.

A comprehensive evaluation of ensemble learning for stock-market prediction

Abstract: Stock-market prediction using machine-learning technique aims at developing effective and efficient models that can provide a better and higher rate of prediction accuracy. Numerous ensemble regressors and classifiers have been applied in stock market predictions, using different combination techniques. However, three precarious issues come in mind when constructing ensemble classifiers and regressors. The first concerns with the choice of base regressor or classifier technique adopted. The second concerns the combination techniques used to assemble multiple regressors or classifiers and the third concerns with the quantum of regressors or classifiers to be ensembled. Subsequently, the number of relevant studies scrutinising these previously mentioned concerns are limited. In this study, we performed an extensive comparative analysis of ensemble techniques such as boosting, bagging, blending and super learners (stacking). Using Decision Trees (DT), Support Vector Machine (SVM) and Neural Network (NN), we constructed twenty-five (25) different ensembled regressors and classifiers. We compared their execution times, accuracy, and error metrics over stock-data from Ghana Stock Exchange (GSE), Johannesburg Stock Exchange (JSE), Bombay Stock Exchange (BSE-SENSEX) and New York Stock Exchange (NYSE), from January 2012 to December 2018. The study outcome shows that stacking and blending ensemble techniques offer higher prediction accuracies (90–100%) and (85.7–100%) respectively, compared with that of bagging (53–97.78%) and boosting (52.7–96.32%). Furthermore, the root means square error (RMSE) recorded by stacking (0.0001–0.001) and blending (0.002–0.01) shows a better fit of ensemble classifiers and regressors based on these two techniques in market analyses compared with bagging (0.01–0.11) and boosting (0.01–0.443). Finally, the results undoubtedly suggest that an innovative study in the domain of stock market direction prediction ought to include ensemble techniques in their sets of algorithms.