Design and Implementation of Hybrid Stock Market Analysis and Prediction using Dynamic Dataset

K M Thanuja1, Dr. N.Deepak Kumar2, Dr.M.Giri3
1P.G Scholar, Department of CSE,Sree Rama Engineering College, Tirupati, Andhra Pradesh, India,
naiduthanuja4@gmail.com
2Professor, Department of Computer Science & Engineering, Sree Rama Engineering College, Tirupati, Andhra Pradesh,
India, deepakkumarsvuphd@gmail.com
3Professor, Department of CSE (AI), Mother Theresa Institute of Engineering and Technology, Palamaner, Andhra
Pradesh, India, dr.m.giri.cse@gmail.com

Abstract - Stock market data is in nonlinear fashion, heavy fluctuated data, time series data, present heavy noisy data, and handling such a data is a one of the crucial tasks. Traditional stock prediction approaches using ML approaches were exhibits moderate results and not appropriate to handle heavy fluctuated and noisy data samples related to stock market. Many of these techniques are depends on short term fluctuations in the stocks andmany be representing wrong stock predictions. In this  research we developed Hybrid Stock Market Analysis andPrediction approach, capable to handle heavy fluctuated  stock market dynamic dataset to enhance prediction, feature encryption techniques are used to convert continuous data samples into consistent representation, and the model can be easily predicting the changes in themarket trend. All the data samples are processed using  hybrid ML approach for modelling based on temporal dependencies and the segments of time are prioritize withhelp of attention technique. Performance of proposed  model is evaluated and compared with LSTM, MLP, RF, SVM, and LR approaches are evaluated in terms of accuracy, precision, recall, F1-score, FPR, FNR, MCC, and  AUC values. Proposed model exhibits greater accuracy, exhibits greater precision, exhibits greater recall, exhibits greater F1-score, exhibits low FPR, exhibits low FNR, exhibits greater MCC, and exhibits greater AUC compare with other similar techniques.Key Words: Machine Learning, Deep Learning, LSTM,Random Forest, Support Vector Machine, MLP, FPR,  FNR, Logistic Regression