Entity

Time filter

Source Type


Singha B.,Camellia Institute of Engineering | Bar N.,University of Calcutta | Das S.K.,University of Calcutta
Journal of Molecular Liquids | Year: 2015

The effects of different operating parameters such as initial pH, initial Pb(II) ion concentration, adsorbent dosages, and contact time are studied to optimize the conditions for maximum batch adsorptive removal of Pb(II) ions from aqueous solution using six different low cost natural bio-sorbents. Applicability of artificial neural network (ANN) analysis is investigated for the removal of Pb(II). Three standard training algorithms (Backpropagation, Levenberg-Marquardt and Scaled Conjugate Gradient) along with four different standard transfer functions in a hidden layer with a linear transfer function in the output layer are used for the analysis. Statistical analysis show that the ANN model with a BP algorithm, using transfer function 1 and 25 processing elements in a single hidden layer gives best predictability of the percentage removal of Pb(II) ions from aqueous solutions. © 2015 Elsevier B.V. All rights reserved. Source


Misra K.,Indian National Institute of Engineering | Chattopadhyay S.,Camellia Institute of Engineering | Kanhar D.,Indian National Institute of Engineering
Journal of Medical Imaging and Health Informatics | Year: 2013

Diagnoses of psychiatric diseases are still a major issue. Two key reasons are-there are variations in the opinions of the medical doctors and the presentation of a disease among the subjects. Given this scenario, the focus of this paper is to develop a hybrid approach for diagnosing adult depression (302 × 16), where each case is represented with 15 symptoms [0, 1] and the corresponding target (that is, grade-probability as 'mild,' 'moderate,' and 'severe'). The proposed hybrid tool is consisted of (a) information gain measure of each symptom/attribute to find the dominant symptoms and (b) a multilayer feedforward-backpropagation neural network, which has been fed with the (i) dominant symptoms and (ii) all symptoms as the inputs to compare its performance. The study observes that with dominant symptoms, the tool is able to classify depression with 98.96% average accuracy, compared to all symptoms, where the average accuracy is 98.91%. The paper concludes that attribute selection procedure based on decision tree learning has increased the efficiency of the tool. © 2013 American Scientific Publishers. Source


Jana D.,Camellia Institute of Engineering | Mandal K.K.,Jadavpur University
Advances in Intelligent Systems and Computing | Year: 2013

This paper presents a novel optimization algorithm for environmentally constrained economic dispatch (ECED) problem using modified real coded genetic algorithm (MRCGA). The ECED problem is formulated as a non-linear constrained multi-objective optimization dilemma satisfying both equality and inequality constraints. The regenerating population procedure is added to the conventional RCGA in order to improve escaping the local minimum solution by a new combination of crossover and mutation technique. To solve ECED problem the predictable RCGA is customized specially by the concept of self adaptation of mutation distribution followed by polynomial mutation approach with arithmetic crossover. To test performance compatibility between them, a six units system is being considered and the better simulation results produce improved solution compare to different methods. © 2013 Springer-Verlag. Source


Mandal K.K.,Jadavpur University | Jana D.,Camellia Institute of Engineering | Tudu B.,Jadavpur University
Advances in Intelligent Systems and Computing | Year: 2013

Optimal reactive power compensation is one of the fundamental issues in the operation of power systems. This paper presents a new improved particle swarm optimization technique called black-hole particle swarm optimization (BHPSO) for optimal reactive power compensation of distribution feeders to avoid premature convergence. The performance of the proposed algorithm is demonstrated on a sample test system. The results obtained by the proposed methods are compared with other methods. The results show that the proposed technique is capable of producing comparable results. © 2013 Springer-Verlag. Source


Chattopadhyay S.,Camellia Institute of Engineering | Rajput S.S.,Indian National Institute of Engineering | Prajesh A.R.,National Institute of Science and Technology
Journal of Medical Imaging and Health Informatics | Year: 2013

Accurate grading of depression is a research challenge. It is important to decide on the treatment plans. This paper proposes a depression classification model using Bayesian approach. In order to achieve this task, 302 real-world depression cases have been collected and used to develop the initial model where the number of symptoms equals to 15 for each case. The load of symptoms is quantified [0, 1] by specialist doctors (psychiatrists) to measure the cumulative grade of the illness, which is labeled as 'Mild/Low,' 'Moderate/Medium,' and 'Severe/High.' In order to reduce the data dimension and extracting significant factors, Paired t-test (PTT), Principal Component Analysis (PCA), Multiple Linear Regressions (MLR), Analysis of Variance (ANOVA) and Factor Analysis (FA) is performed. All these techniques are tried because we have no prior idea which would effectively serve the purpose. After rigorous experiments, it is noted that 4 symptoms are found significant (p value < 0.05) using PTT. It reduces the data matrix from 302 x 15 to 302 x 4. It helps reducing the complexity in the intended classifier design. At first a Naïve Baye's classifier is designed. It can only grade depression with 19% accuracy. Hence, it is further trained using Hill Climbing Search (HCS) algorithm to develop a Learned Bayesian Classifier. Throughout the design and developmental processes, 70% of the data has been used for training and the rest tests the classifier's performance. Experimental results show that the average accuracy achieved by the learned classifier is close to 80%. Copyright © 2013 American Scientific Publishers All rights reserved. Source

Discover hidden collaborations