About Me |
---|
Sasmita Kumari Nayak is currently working as an Associate Professor at Centurion University of Technology and Management, Bhubaneswar, Odisha. |
30 Research Paper Publications, 5 Conference Proceedings, two Book Chapters, and 4 Patents (2 granted and 2 published).
Sl. No. | Title | Issuer |
---|
This paper deals with the problem of dynamic task scheduling in grid environment of multi-processors. First, this paper formulates task scheduling as an optimization problem and then optimizes with a novel hybrid optimization algorithm. The proposed algorithm combines the merits of Genetic Algorithm and Bacteria Foraging optimization. The simulation result proves the superior performance with the proposed
algorithm.
The multiprocessor scheduling problem consists of a set of tasks to be performed using a finite number of processors. This paper deals with the problem in a heterogeneous processing environment. A nature-inspired meta-heuristic algorithm, Water Cycle Algorithm (WCA) is being used for the purpose. For the purpose of comparison, contemporary strategies using Genetic Algorithm (GA), Bacteria Foraging Optimization (BFO) and Genetic based Bacteria Foraging (GBF) found in the literature also reproduced in this paper. Because of close relationships between the matrixes formed by the problem with those of the WCA, proposed strategy of scheduling outperforms GA and GBF based strategies.
The multiprocessor scheduling problem consists of a set of tasks to be performed using a finite number of processors. This paper deals with the problem in a heterogeneous processing environment. A nature-inspired metaheuristic algorithm, water cycle algorithm (WCA), is being used for the purpose. For the purpose of comparison, contemporary strategies using genetic algorithm (GA), bacteria foraging optimization (BFO) and genetic-based bacteria foraging (GBF) found in the literature also reproduced in this paper. Because of close relationships between the matrixes formed by the problem with those of the WCA, proposed strategy of
scheduling outperforms GA- and GBF-based strategies.
This paper manages the issue of Multiprocessor scheduling Problem is one of the most challenging problems in distributed computing system. Many researchers solved the multiprocessor scheduling problem as static. But in this paper uses the dynamic multiprocessor scheduling problem which is an advanced area. Dynamic allocation strategies can be connected to huge arrangements of genuine applications that can be planned in a way that takes into account deterministic execution. In the first place, here defines the Dynamic Multiprocessor scheduling, which is an optimization problem, after that it optimizes the execution time of various tasks assigned to the processors with a Krill Herd Algorithm (KHA). In recent times, a robust meta-heuristic optimization algorithm, known as Krill Herd, which is used for global optimization to enhance the execution of the multiprocessor scheduling problem but other traditional algorithms stuck in local optimization. In this paper with the end goal of comparison, contemporary methodologies utilizing Genetic Algorithm (GA), Bacteria Foraging Optimization (BFO) and Genetic based Bacteria Foraging (GBF) found in the literature. Here, it demonstrates the better performance of Krill Herd Algorithm with the above mentioned methods by simulation process.
Machine learning and Data Mining Techniques are mainly used for many Real world problems. The traditional methods include training the data and test .But it will not be applicable for real world scenario. Some of the reason may be the cost of training data and inability to get those. These drawbacks are giving rise to the concept known as Transfer Learning.It ensures that training data must be independent and distributed identically.Transfer Learning is considered as a solution to the insufficient training data.
The multiple processor scheduling problem characterizes that different processor comprises of an arrangement of jobs or tasks designate proficient utilizing a limited number of processors. Herein development a multi-objective algorithm utilizing Symbiotic Organisms Search algorithm (SOSA) for scheduling an arrangement of reliant on tasks on obtainable resources in a multiple processor environment which minimizes the execution time and maximize the processor utilization. SOSA is a nature-inspired meta-heuristic algorithm utilized to compare with other meta-heuristic algorithms such as Water cycle algorithm (WCA), Genetic algorithm based Bacteria foraging optimization (GBF), Bacteria Foraging Optimization (BFO) and Genetic Algorithm (GA). SOSA reproduces the advantageous association methodologies received by life forms to survive and engender in the biological-community (ecosystem). Based on experimental results, we find the execution time as well as processor utilization using SOSA technique and then compare with the other mentioned algorithms. Acquired outcomes affirm the incredible execution of the SOSA in solving the multiple processor scheduling problems.
At the present time, in contrast with the past, this paper achieves the most challenging problems in distributed computing system named as Dynamic Task Scheduling Problem (DTSP). In this paper, the DTSP optimizes with a multi-objective meta-heuristics. At this point, the Grey Wolf Optimization Algorithm (GWOA) is the use of global optimization to augment the execution of the Dynamic task scheduling problem. By simulation process, the GWOA provides better performance by comparing the results of other optimization algorithms.
During the beginning of seventieth centuries, human facial recognition has become one among the researched areas in the area of finger print scanning and computer vision. Identifying a person with an image has been popularized through the mass media. The recent technologies are totally focusing on developing the smart systems that will recognize the faces for biometric purposes. In this context automatic face recognition is applied for security purposes to find the criminal, attendance system, scientific laboratories etc. This research paper presents the frame work for real time face detection. However, it is less robust to finger print or retina scanning. This paper describes about the face detection and recognition. These technologies are available in the Open-Computer-Vision (OpenCV) library and methodology to implement them using Python in image processing and machine learning. For face detection, Haar-Cascades algorithms were used and for face recognition the algorithm like Eigen faces, and Local binary pattern histograms were used.
This paper aims at combining real-time object detection and recognition with suitable deep learning methods in order to detect and recognize objects position as well as the names of multiple objects detected by the camera using an object detector algorithm. This is to aid the visually impaired user without the help of any other person. The image and video processing algorithms were designed to take real-time inputs from the camera, Deep Neural Networks were used to predict the objects and uses Google’s famous Text-To-Speech (gTTS) API module for the anticipated voice output precisely detecting and recognizing the category or class of objects and locations contained. Our best result shows that the system recognizes 91 categories of outdoor objects and produces the output in speech i.e. in an audio format even when a reduced amount of spectral information from the data is available.
Face of a human being is an important part. One person can be differentiated from the other with the help of it due to some unique features associated. Biometrics is one of the important area, where Facial Recognition plays a major part. Scope of this paper is to review on facial recognition. It has the objectivity to find and survey different Facial recognition techniques. These are the real world applicabilities in the area of Artificial Intelligence and Machine Learning. This acts as a frame work for real time face recognition and detection. Face Detection is regarded as a technology which considers the location and size of a face in digital image. This Paper is intended for presenting a comprehensive review of Facial Recognition Techniques.
Hepatitis is considered as one among the most dangerous diseases in the world. The inflammation situation of liver is referred to as a hepatitis disease. Generally, the viral infection is the reason of this disease. In this world, the serious issue of health sectors is hepatitis disease due to the late diagnosis as well as less number of casual agents. To overcome this, quick diagnosis is very much essential. Data mining is a new branch of science, which helps medical doctors in suitable decision making. In data mining, machine learning algorithm and reduction features are helpful to decrease the method of disease diagnosis and complexity of the issue, respectively. Nowadays, healthcare industry is rich in data with poor knowledge. The techniques of data mining are utilized for extraction of knowledge from the data and take the proper decision making for diagnosis. Increasing research on hepatitis disease predicting systems, it become significant to summarize the completely by making research on it. The main purpose is to summarize the recent research with comparative outcomes, which has been done on hepatitis disease prediction and also make analytical conclusions. This paper uses and summarises data mining and machine learning techniques along with their complexities.
Weather forecasts can be performed by collecting large amount of data about the present condition of the environment. The state of atmosphere or environment can be the temperature, humidity and wind.
The future can be determined by the atmosphere evolution and this evolution can be done through
meteorology with understanding of atmospherical processes. The foundation stone of Indian economy is
the agriculture, which depends on the rain. Hence, early prediction of rain is highly necessary in India for
agriculture. In worldwide, rainfall prediction is one among the challenging issues. Here, varied and most
popular machine learning models are utilized for forecasting the rainfall. The outcome from the model
gives the centre of the weather forecasts for predicting that whether tomorrow will rain or not. The
experimental results will compare and show that good level of accuracy with the help of machine learning
algorithms.
Electricity Consumption is one among widely studied section of computer architecture
for more than decades. Electricity Adoption is one of the parameter in Machine Learning. It is
one of the emerging field in the research. It keeps an eye on high accuracy without any kind of
computation constraint. The paper has the objectivity to analyze Machine Learning algorithms
effectiveness, after being applied to the electricity consumption prediction. Load management
and demand response, high dimensional data sets are much effective variable selection, accurate
prediction for electricity market pricing. It retains economic mechanisms to the largest. Our goal
is about important guidelines offering to the Machine Learning community and provide basic
knowledge of building specific electricity consumption estimation methods for machine learning
algorithms. This Paper reviews about the Conventional Machine Learning Models as well as the
recent models, allowing predicting electricity consumption. A number of research works are
concerned with the set of structural models and its corresponding applicabilities are introduced.
The predictions are proposed for the research reference, depending upon previous work analogy.
For farmers, India is very famous in agriculture. The economic growths for all nations are dependent on agricultural products. Due to plant diseases, the quantity and quality of agriculture yield are reduced. The study of
plant disease refers to the study of clearly visible patterns of plant leaves. So, recognition of the unhealthy regions
of plants may be thought about the way of saving the decrease of productivity and crops. The early-stage diagnosis
of plant diseases like viruses, bacteria, fungi, etc. is most essential to control and cure the disease. The manual
identifications of diseases are a time-consuming process. Hence, some experts are required to recognize the
disease. There are varied standard methods such as classification model, image processing, and machine learning
models that are utilized to detect and recognize the disease on agricultural yields. This article provides varied
existing models are made familiar with the detection of disease in agricultural product. Further, it presents a survey
on varied classification models with the analysis, which could be utilized to classify and identify the plant leaf
diseases. It also discusses the outline of segmentation, feature extraction, and varied classifier techniques.
In analysis of crop yield production, an emerging research field is the Data Mining. Crop yield is a highly
complex trait in agriculture. Basically, data mining is a method for analysing data from varied viewpoints and
summarized the same into important information. For crop yield prediction, Machine learning is also a
significant decision support tool that includes supporting decisions upon, which crops to cultivate and what
actions should be taken while growing season of the yields. The results of the prediction will be made available
to the farmer. For the research of crop yield prediction, various machine learning models have been employed.
In this article, the prediction has been done for coconut crop. This paper applied four different supervised
techniques like Random Forest, Gradient Boosting, Support Vector Machine, and Decision Tree Regression
techniques to get the accuracy of coconut crop yield. This study proposed and implemented all the models to
predict coconut crop by using the previous data. The outcomes of simulation illustrates that the proposed work
efficiently for prediction of coconut crop.
In analysis of crop yield production, emerging research fields are Machine Learning Model and Data Mining. In
agriculture, production of crops is a very much complex issue. Also, it is a big issue for farmers. Analysis of
crop yield production is a highly essential step for predicting the production of crop. This is an initial step of this
issue as well as for all types of problems. The analysis step is used to analyse for cultivation of which crops and
what actions should be taken whilegrowing season of the yields, which helps to the farmer for production of
crop. In this article, the analysis has been done for crops in terms of different zones, state wise top highest crops
and their comparisons. The outcomes of simulation illustrate the prediction and cultivation of crops that helps to
farmer for cultivating which crop in which region of India, so that the farmers will be benefited
The stock markets contribute a largescope in economic development of India. The banking industrygrip majority
share between other industries in Indian stock trading consequence. The investors in the stock market use to
bear certain risk for their predictable returns in the future. Investment decisions are usually taken by considering
different fundamental factors both internal and external. Apart from fundamental factors which replicated in the
security prices, there are numerous additional factors that can influence investment are stock prices, volume of
trading, spread, turnover etc.The paper explores the effect of different variables on the high stock price of Visa
Steelconsidering daily data over the period 4 Jan 2010 to 23 Apr 2020. For the study the “weighted average
price (WAP), number of shares, number of trades, total turnover (in INR), deliverable quantity, percent
deliverable quantity to traded quantity, spread high and low, spread open and close and the high stock price of
the organizationâ€are noted. High stock price is considered as output while other parameters are used as input.
Pipeline Pilot module of Biovia software (DassaultSystems of France) is used for analysis. The software
provides different built-in components to develop a machine learning model and use the model for prediction.
This Machine Learning article is used to predict or detect the disease based on the symptoms given by the user. For small problems, the users have to go personally to the hospital for check-up which is more time consuming. Also handling the telephonic calls for appointments is quite hectic. Such problems can be solved by using disease prediction application. Over the years, the use of the specific disease prediction tools has been increased due to a variety of diseases and less doctor-patient ratio. Thus, in this system, we are concentrating on providing immediate and accurate disease prediction to the users about the symptoms they enter along with the severity of disease predicted. Best suitable algorithm and doctor consultation will be given in this project. For prediction of diseases, different machine learning algorithms are used to ensure quick and accurate predictions. Disease Prediction is done by implementing four different machine learning algorithms such as Naive Bayes Classifier, Random Forest Classifier, Decision Tree Classifier and K-nearest neighbour Classifier. This Classifiers calculates the probability of the disease. Therefore, average prediction accuracy probability 90% is obtained. “Disease Prediction†system based on predictive modeling predicts the disease of the user on the basis of the symptoms that user provides as an input to the system. The system analyzes the symptoms provided by the user as input and gives the probability of the disease as an output. With big data growth in biomedical and healthcare communities, accurate analysis of medical data benefits early disease detection and patient care.
Rainfall prediction is important because heavy rainfall can lead to many disasters. The prediction helps people to take preventive measures and moreover the prediction should be accurate. Heavy precipitation prediction could be a major drawback of earth science department because it is closely related to the economy and lifetime of human. Accuracy of rainfall statement has nice importance for countries like India whose economy is basically dependent on agriculture. The dynamic nature of atmosphere, applied mathematics technique fails to provide sensible accuracy of precipitation statement. Intention of this project is to offer non-experts easy access to the techniques, approaches utilized in the sector of precipitation prediction and provide a comparative study among the various machine learning techniques.
Owing to its recent advance, machine learning has spawned a large collection of solar forecasting works. In particular, machine learning is currently one of the most popular approaches for hourly solar forecasting. Nevertheless, there is evidently a myth on forecast accuracy-virtually all research papers claim superiority over others. Apparently, the "best" model can only be selected with hindsight, i.e., after empirical evaluation. For any new forecasting project, it is irrational for solar forecasters to bet on a single model from the start. To ensure a fair comparison, no hybrid model is considered, and only off-the-shelf implementations of these algorithms are used. Moreover, all models are trained using the automatic tuning algorithm available in the package. It is found that tree-based methods consistently perform well in terms of overall results. Increasing penetration of distributed renewable power means that reliable generation forecasts are required for grid operation. The present work aims at combining state of the art implementations of the Weather Research and Forecasting (WRF) model with multivariate statistical learning techniques to provide the most accurate forecasts of day-ahead hourly irradiance.
The Kepler exoplanet mission is specially organised for searching of the Milky Way galaxy to discover many earth-size and tiny planets near or in the habitable zone. Exoplanet means orbiting of planet around the other stars beyond our solar system. The dataset contains 9564 samples and 50 columns which is collected from kaggle website. The dataset, target variable is KOI-disposition which contains confirmed, false positive and candidate. Out of 9564 samples we found 5000 samples are false positive, each confirmed and candidate are 2282. In the Milky Way galaxy many stars and planets are there, but we have considered some of them. We have used machine learning algorithms like decision tree, random forest, KNN classification and Naive Bayes classification on stars and planets for searching of exoplanets beyond our stellar atmosphere.
Food processing businesses play a crucial role in ensuring that people have access to
nutritious food. These companies can either produce food directly for consumers or provide
ingredients for other businesses to use in their products. Many food processing firms rely on
interacting directly with customers to sell their products. However, this can be a challenge
during the COVID-19 pandemic and other situations that limit face-to-face interactions. One
potential solution to this issue is the use of computer vision to analyze food quality in an
automated, non-invasive, and cost-effective way. This type of technology can help businesses
ensure that they are providing high-quality products to their customers. A study has presented
an OpenCV Python library-based system that uses the Convolution Neural Network (CNN) to
identify and assess food quality.
Food processing businesses play a crucial role in ensuring that people have access to
nutritious food. These companies can either produce food directly for consumers or provide
ingredients for other businesses to use in their products. Many food processing firms rely on
interacting directly with customers to sell their products. However, this can be a challenge
during the COVID-19 pandemic and other situations that limit face-to-face interactions. One
potential solution to this issue is the use of computer vision to analyze food quality in an
automated, non-invasive, and cost-effective way. This type of technology can help businesses
ensure that they are providing high-quality products to their customers. A study has presented
an OpenCV Python library-based system that uses the Convolution Neural Network (CNN) to
identify and assess food quality.