Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
6 results
Search Results
Item End-to-end monitoring of cloud resources using trust(Springer Verlag, 2019) Divakarla, D.; Chandrasekaran, K.Cloud computing is the fast-growing technology in the current era. With the growth, it brings with it multiple issues like privacy, security. Trusting the service provider and his resources is the major drawback for the user. In this paper, we have proposed a trust model which builds an end-to-end trust between the service provider resources and user. Our model calculates trust based on four parameters, namely utilization, saturation, failure, and availability. Our model builds a strong trust decision which enables the user to trust the resources of the service provider. © Springer Nature Singapore Pte Ltd. 2019.Item Role of Activation Functions and Order of Input Sequences in Question Answering(Springer, 2020) Chenna Keshava, B.S.; Sumukha, P.K.; Chandrasekaran, K.; Divakarla, D.This paper describes a solution for the Question Answering problem in Natural Language Processing using LSTMs. We perform an analysis on the effect of choice of activation functions in the final layer of LSTM cell on the accuracy. Facebook Research’s bAbI dataset is used for our experiments. We also propose an alternative solution, which exploits the language structure and order of words in the English language, i.e. reversing the order of paragraph will introduce many short-term dependencies between the textual data and the initial tokens of a question. This method improves the accuracy in more than half of the tasks by more than 30% over the current state of the art. Our contributions in this paper are improving the accuracy of most of the Q&A tasks by reversing the order of words in the query and the story sections. Also, we have provided a comparison of different activation functions and their respective accuracies with respect to all the 20 different NLP tasks. © 2020, Springer Nature Singapore Pte Ltd.Item Optimized diet plan using unbounded knapsack Algorithm(Institute of Electrical and Electronics Engineers Inc., 2020) Bobade, P.; Kumar, P.; Chandrasekaran, K.; Divakarla, D.Cholesterol, hypertension and diabetes are the three major chronic diseases from which most of the people suffers and these peoples often use search engines to acquire related information about these problems. But, almost every information related to diet on the internet isn't suitable for people to gather information about the diet suggestions. A system for diet suggestion which can advocate a prudent diet for such peoples is suggested in this paper. We designed a system that recommends a proper diet which has the adequate knowledge of three above mentioned highly chronic diseases. We propose a solution to the menu recommending problem using the optimization algorithm known as unbounded knapsack. We designed a model which satisfies the nutritional requirements of individuals while imposing the 'Laws of Nutrition', a set of hypothesis used by almost all Latin America's nutrition scientists. This prototype corresponds to a numerical optimization problem with constraints. We design a menu items generator application model to set up a convenient menu for a user with different properties. © 2020 IEEE.Item Towards a Federated Learning Approach for NLP Applications(Springer Science and Business Media Deutschland GmbH, 2021) Prabhu, O.S.; Gupta, P.K.; Shashank, P.; Chandrasekaran, K.; Divakarla, D.Traditional machine learning involves the collection of training data to a centralized location. This collected data is prone to misuse and data breach. Federated learning is a promising solution for reducing the possibility of misusing sensitive user data in machine learning systems. In recent years, there has been an increase in the adoption of federated learning in healthcare applications. On the other hand, personal data such as text messages and emails also contain highly sensitive data, typically used in natural language processing (NLP) applications. In this paper, we investigate the adoption of federated learning approach in the domain of NLP requiring sensitive data. For this purpose, we have developed a federated learning infrastructure that performs training on remote devices without the need to share data. We demonstrate the usability of this infrastructure for NLP by focusing on sentiment analysis. The results show that the federated learning approach trained a model with comparable test accuracy to the centralized approach. Therefore, federated learning is a viable alternative for developing NLP models to preserve the privacy of data. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item Fully Decentralized Blockchain and Browser-Based Volunteer Computing Platform(Springer Science and Business Media Deutschland GmbH, 2022) Bharadwaj, K.S.S.; Dharanikota, S.; Honawad, A.; Divakarla, D.; Chandrasekaran, K.Volunteer computing allows individuals, who have access to computing resources that are currently idle, to allocate them to perform useful work. This paradigm has existed since a long time and is evolving by the day with the advent of novel approaches such as browser-based volunteer computing. But most of these solutions have a degree of centralization in their architecture and are prone to single point of failure issues, or require explicit trust in the entities that manage the network. This paper proposes the use of blockchain to eliminate these drawbacks of traditional volunteer computing platforms, at the same time preserving the ability to make the architecture entirely browser-based. The key focus of our proposal is on resilience. Resilience is achieved by making use of the decentralized storage system, InterPlanetary File System. Finally, we present a prototype implementation of our ideas. We evaluate our system by solving NP-problems using the prototype. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item A Machine Learning Approach for Load Balancing in a Multi-cloud Environment(Springer Science and Business Media Deutschland GmbH, 2022) Divakarla, D.; Chandrasekaran, K.A multi-cloud environment makes use of two or more cloud computing services from different cloud vendors. A typical multi-cloud environment can consist of either only private clouds or only public clouds or a combination of both. Load balancing mechanism is essential in such a computing environment to distribute user requests or network load efficiently across multiple servers or virtual machines, ensuring high availability and reliability. Scalability is also achieved by sending requests only to those servers that are healthy and available to take up the computing workload and thus providing the flexibility to scale up and scale down to satisfy QoS requirements as well, in order to save costs. In our proposed model, a time series-based approach as well as predictive load balancing has been experimented and the results are presented. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.