Papers by Rushali Deshmukh
PLOS ONE, 2020
The literature provides strong evidence that stock price values can be predicted from past price ... more The literature provides strong evidence that stock price values can be predicted from past price data. Principal component analysis (PCA) identifies a small number of principle components that explain most of the variation in a data set. This method is often used for dimensionality reduction and analysis of the data. In this paper, we develop a general method for stock price prediction using time-varying covariance information. To address the time-varying nature of financial time series, we assign exponential weights to the price data so that recent data points are weighted more heavily. Our proposed method involves a dimension-reduction operation constructed based on principle components. Projecting the noisy observation onto a principle subspace results in a well-conditioned problem. We illustrate our results based on historical daily price data for 150 companies from different market-capitalization categories. We compare the performance of our method to two other methods: Gauss-Bayes, which is numerically demanding, and moving average, a simple method often used by technical traders and researchers. We investigate the results based on mean squared error and directional change statistic of prediction, as measures of performance, and volatility of prediction as a measure of risk.
2023 International Conference on Sustainable Computing and Data Communication Systems (ICSCDS)
International Journal of Recent Technology and Engineering (IJRTE), 2019
Poem a piece of writing in which the expression of feelings and ideas is given intensity by parti... more Poem a piece of writing in which the expression of feelings and ideas is given intensity by particular attention to diction (sometimes involving rhyme), rhythm, and imagery. It is used for showing different views. Every poet writes a poem with a different intention and different views. In the proposed system we have classified the poem according to its sentiments by using words of different categories. Machine learning algorithm SVM classifier is used for differencing the class of the poem. This system also enables the user to search the poem based on the poet name and poet type. For 341 poems of five categories 'Friend', 'Prem', 'Bhakti', 'Prerna' and 'Desh' accuracy achieved is 93.54%.
Turkish Journal of Computer and Mathematics Education (TURCOMAT)
Social media applications like Twitter, Instagram, Facebook have helped people to connect to each... more Social media applications like Twitter, Instagram, Facebook have helped people to connect to each other. This has been eased due to high-speed internet. However, this has invited various spam messages through tweets or Facebook. The sole purpose of such messages is aggregation or exploitation of personal data in terms of finances or medical records, political benefit’s or community violence. This makes spam detection an extreme value-added service. We tend to recommend a 1D CNN algorithmic technique and compare results with variants of CNN and with boosting algorithms. The model is braced with linguistics data in the illustration of the words with the assistance of knowledge-bases such as Word2vec and fast ext. This improves the end to end performance, by providing higher linguistics vector illustration of input testing words. Projected Experimental results show the efficiency of the projected approach from the point of view of accuracy, F1-score and response time.
Turkish Journal of Computer and Mathematics Education (TURCOMAT)
People have a tendency to analyze existing strategies and so planned new strategies for inventory... more People have a tendency to analyze existing strategies and so planned new strategies for inventory prediction. We have used Sentiment evaluation and Technical evaluation through NLP and Deep mastering approach. In order to exploit benefits of sentiment analysis on enterprise associated inventory, we have proposed a model that will use the sentiment analysis on twits associated with special sectors that are Information Technology sector, Banking sector, Pharmaceutical sector, Automobile sector, Infrastructure sector which are extracted from twitter. These twits are extracted from twitter for calculating polarity. The rating of sentiment analysis is calculated here by using Natural Language Processing’s method. According to sector we've taken five groups. Top four performer businesses of every sector. Using polarity score we got finalized pinnacle ten groups with great sentiment rating. We then downloaded the CSV facts of historical share charge of top ten organizations that we'...
These days dependent on accessible clinical examination specialist is utilizing experimentation a... more These days dependent on accessible clinical examination specialist is utilizing experimentation approach for anticipating infections. To foresee the infection is one of the real difficulties in past years and today too. Among the different maladies, heart disease is normal nowadays due to pushing the issue and remaining burden. On the basis of available symptoms and patients health, there is a great need of some system that predicts the diseases at an early stage. This research paper provides a survey of different data mining techniques which have already used for heart disease prediction and understand how efficiently Electronic Health Record (EHRs) offer clinicians functions that could never be achieved with written records. We are proposing the Disease status identification system by the Machine Learning algorithm. By using this algorithm we are going to predict Heart Disease based on their symptoms.
Advances in Intelligent Systems and Computing, 2021
Easy access of information and data on the internet enable the users download and store lots of d... more Easy access of information and data on the internet enable the users download and store lots of data in their system. Most of storage space is occupied by the duplicate data that have been downloaded or gathered from other resources and stored in the system thus increasing the requirement of the storage space. While taking the backup to the cloud same redundant data is uploaded to the cloud storage. With increase in the amount of such redundant data proper utilization of the storage resources and bandwidth is not possible. Once the data is uploaded at the cloud the user’s are not sure how secure their data is. In this paper block level deduplication approach is applied to reduce the data redundancy. In order to maintain the secrecy of the data SHA-256 a cryptographic hash algorithm is implemented. Also compression is performed on the data so that proper bandwidth utilization is possible. Keywords— Data deduplication; Data compression; Indexing; cloud; SHA-256; Collision resistant. —...
2020 2nd International Conference on Innovative Mechanisms for Industry Applications (ICIMIA), 2020
For many natural language processing applications, part of speech (POS) tagging remains as a prel... more For many natural language processing applications, part of speech (POS) tagging remains as a preliminary task. Marathi, is observed as a popular language in India but it only has limited tools and corpus for NLP applications. An accurate POS tagger is essential for many NLP tasks like sentiment analysis, named entity recognition, dependency parsing, etc. This research work proposes a deep learning model and bidirectional long short-term memory (Bi-LSTM) model to perform POS tagging for Marathi text. We achieved an accuracy of 85% for the deep learning model and 97% for the Bi-LSTM model. Our contribution here is based on three folds - building a deep learning model, building the Bi-LSTM model, comparison with machine learning techniques for the same dataset.
International Journal of Engineering and Advanced Technology, 2019
The classification technique is most important for supervised and semi supervised base machine le... more The classification technique is most important for supervised and semi supervised base machine learning task. Many classification algorithms has introduced already for existing systems. Class-label classification is an important machine learning task wherein one assigns a subset of candidate without label to an object. Classification of various document models based on short text, metadata, heading levels these are the existing techniques which are introduced in literature survey. Sometime whole data reading and processing might be take a much time for classification, so it increase the time complexity for entire system. We proposed a new document classification method based on deep learning using NLP and machine learning approach. In this work system has several attractive properties: it captures some metadata from entire abstract section and built the training set first. Once complete all document process, it deals with optimization algorithm. Recurrent Neural Network has used to ...
International Journal of Computer Sciences and Engineering, 2018
2015 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT), 2015
Mining association rules is a fundamental data mining task. Association rule greatly help to iden... more Mining association rules is a fundamental data mining task. Association rule greatly help to identify trends and pattern from huge data set. Algorithms for mining association rules put more stress on positive rules rather than negative rules. Negative rules specify the attribute present in the data set to the attribute absent. In this paper we propose an algorithm BMPNAR, Best M Positive Negative Association Rules Algorithm, in order to get a reduced set of limited number of association rules which are then classified using Firefly algorithm. The algorithm BMPNAR is an extension to MOPNAR algorithm. It is a combination of MOPNAR and Topk algorithm. We let the user specify the number of rules to be generated. It gives us ranked association rules. These rules are then classified by applying FireFly algorithm for analysis purpose. The system designed is supposed to generate best M classified rules. The dataset used is Keel Dataset. We give the comparative study of previous and new algorithm in terms of execution time and space required.
Uploads
Papers by Rushali Deshmukh