OptiSense: A Hybrid Deep Learning System for NSE Option Strategy Prediction and Analysis

Sanika S. Todkari, Anujkumar V. Kate and Tejas V. Joshi

International Journal of Scientific Research in Science and Technology, 2026, vol. 13, issue 3, 1059-1069

Abstract: In the National Stock Exchange (NSE), options trading is highly nonlinear and is affected by the implied volatility, the sensitivity of the strike price, open interest and time decay. The interactions are complex, and traditional single-model forecasting systems are not well suited to capture these interactions. In this paper, OptiSense, a hybrid intelligent framework for predicting and analyzing NSE option strategy using deep learning and ensemble learning methods is presented. The proposed system is an ensemble aggregation of Convolutional Feed-Forward Network (Conv-FFN), Transformer Encoder-Decoder and BiLSTM with Self-Attention architectures. An intelligent routing mechanism dynamically routes CE/PE option contracts to a specialized model based on BiLSTM and the general market instruments are passed through the ensemble architecture. The framework also includes uncertainty estimation (Monte Carlo Dropout) and real-time safety filters for higher trading reliability. The system uses thirty one domain-specific features from the NSE option chain data obtained with the help of Zerodha kite API such as RSI, Bollinger Bands, Open Interest, signed moneyness, proxies of implied volatility, etc. The experimental evaluation using 315,166 samples from 50 NSE instruments shows that the Transformer Encoder-Decoder model has an accuracy of 86.7% ± 0.6%, Conv-FFN model has an accuracy of 80.5% ± 0.7%, and the simple probability averaging model has the highest accuracy of 87.7%. With proper anti-leakage validation, it is possible to achieve an accuracy of around 60% for options-specific prediction, which is close to the actual level of difficulty in predicting the price of derivatives.

Keywords: Option Chain Prediction; Deep Learning; Ensemble Learning; NSE; Transformer; BiLSTM; XGBoost; Financial Forecasting; Uncertainty Estimation (search for similar items in EconPapers)
Date: 2026
References: Add references at CitEc
Citations:

Downloads: (external link)
https://ijsrst.com/home/article/view/IJSRST261332045 Abstract page (text/html)
https://ijsrst.com/home/article/download/IJSRST261332045/IJSRST261332045 Full text (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:etm:ijsrst:v13:y2026:i3:id:1698

DOI: 10.32628/IJSRST261332045

Access Statistics for this article

More articles in International Journal of Scientific Research in Science and Technology from Technoscience Academy
Bibliographic data for series maintained by Pankaj Sharma ().