Towards A Deep FLANN For Prediction Of Stock Market Returns

Gaurav, Prashant; Singhal, Abhishek; Mani, Ashish

doi:10.1109/rdcape47089.2019.8979088

Cited by 1 publication

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Efficient functioning of stock markets requires market intermediaries who trade stocks for a short duration and keep the market liquid. Machine learning algorithms have been proposed to help such market intermediaries make better predictions for the shortterm price movements [2,6,7]. Both [6] and [7] use LSTMs to predict the stock price.…”

Section: Introduction and Related Workmentioning

confidence: 99%

“…In particular, [7] used sparse auto-encoders with 1-D residual convolutional networks to denoise the data and improve the mean absolute percentage error (MAPE), while [6] uses an Attention [1] mechanism to improve the mean squared error (MSE) in stock price prediction. In [2], they use a deep FLANN (functional link artificial neural network) architecture, which is similar to a feed-forward Neural Network (NN) with time-varying weights to predict the stock prices.…”

Section: Introduction and Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Deep Stock Predictions

Doshi¹,

Alexander²,

Sachdeva³

et al. 2020

Preprint

View full text Add to dashboard Cite

Forecasting stock prices can be interpreted as a time series prediction problem, for which Long Short Term Memory (LSTM) neural networks are often used due to their architecture specifically built to solve such problems. In this paper, we consider the design of a trading strategy that performs portfolio optimization using the LSTM stock price prediction for four different companies. We then customize the loss function used to train the LSTM to increase the profit earned. Moreover, we propose a data driven approach for optimal selection of window length and multi-step prediction length, and consider the addition of analyst calls as technical indicators to a multi-stack Bidirectional LSTM strengthened by the addition of Attention units. We find the LSTM model with the customized loss function to have an improved performance in the training bot over a regressive baseline such as ARIMA, while the addition of analyst call does improve the performance for certain datasets.

show abstract

Section: Introduction and Related Workmentioning

confidence: 99%