Bøker av Chandan Srivastava

This book presents an overview of the evolution and opportunities associated with traditional as well as upcoming fields in the areas of materials, metallurgy, and manufacturing. There are a lot of interesting fields at this trijunction, such as alloy design, bio-materials, composites, high entropy alloys, sensors, electronic materials, and materials degradation. The progress in these fields is further fuelled by the advances in the analysis and fabrication techniques such as correlative microscopy, additive manufacturing, and surface engineering. This book discusses the above topics/fields covering advanced analysis techniques, fabrication methods, and various technological applications. Every chapter walks through the basics of the respective field and comprehensively discusses the current developments and future avenues, to arrive at a point where the reader acquires an overall view of the field. Special emphasis is given to the scientific fundamentals and application potential, in a way that readers of all backgrounds can get benefited. The chapters connect the current developments with the future avenues, to help the researchers foresee the future technologies, in their respective fields. This text will appeal to experienced researchers, practitioners, and students alike. 

This book is based on the basis that clustering and neural networks methods. The clustering algorithms (k-means, and k-medoids) are describe the analysis on noise data i.e. preclassification for robust model development. The better choice of cluster are forming by Euclidean statistical clustering algorithms, are able to preclassified data into significant groups. We assume that both methods are better predicted on different example in real analysis. The recurrent backpropagation is one of the best optimization techniques for minimizing the error and achieve the best optimal result. Since we have input unit, output unit, and eventually hidden unit; we could say that this is supervised optimization learning process. The optimization process to minimizing the error and get the activated network till that all weight of the network are going to reach equilibrium state). This process usually take more time because every output of network add-up with input again and train network (weight) to reach the equilibrium state (optimal solution).Reported results and graphical user interface (GUI) snapshot, showing algorithms are integrated well in software package (simulator).

This book is based on the basis that machine learningalgorithms and rank based statistical methods are abetter choice to develop a robust model in logicalsituations. We designed experimental setup for datacollection, developed unique class of model includingvariable selection, and detection methods. Theselected significant variables provide a unique classof model for all six participants. We emphasize thebest selected variables have good information formodel development, and each selected variable have noerror i.e.; AUC=1, with forward selection and supportvector data description classifier. Basically, wedeveloped a unique class of model using six differentclasses of subjects, predicting elderly fallprevention, and after doing external validation withseventh class of subject, we reached a uniquesolution. Sections one is research introduction,section two is all about research design and dataanalysis, section three and four give extensivedevelopment of model for variable selection and oneclass classifier. Then finally given the conclusionand future aspect of whole study.