A stock market is a compilation of buyers and sellers of stock which represent ownership claims on businesses; these may include securities listed on a public stock.
And the stock price is the price of a single share of a number of saleable stocks of a company, derivative or other financial asset.
By knowing the stock price of a particular company we can known the companies current value or its market value, so the price represents how much the stock trades at or what is the price agreed upon by a buyer or a seller.
Prediction of stock price through deep learning is a very challenging task because stock prices are full of uncertainties .
1)This work of stock predicting uses, sparse autoencoders with one-dimension (1-D) residual convolutional networks which is a deep learning models to predict the data. Than (LSTM) or Long Short Term Memory is used to indicate the price.
2)In past the features like indices, price and macroeconomic variables were used to predict the next days price.
3)Many experiments result shows that 1-D residual convolutional networks can de-noise data and extract deep features better than a model that combines (WT) or wavelet transforms and (SAEs) or stacked autoencoders .
4)In addition, we compare the performances of model with two different forecast targets of stock price:
a) Absolute stock price: Absolute value refers to a business valuation method that uses discounted cash flow analysis to find a company’s financial worth. Investors can find if a stock is currently under or overvalued by comparing what a company’s share price should be given its absolute value to the stock’s current price.
b) Price rate of change: The value of the dollar is both caused and reflected by interest rates, and interest rates have much to do with stock prices. Therefore, exchange rates affect stock prices and can be used to make predictions about the market.
The results show that predicting stock price through price rate of change is better than predicting absolute prices directly.