See attach Journal Article and write 3-4 page review. instru…

Title: A Literature Review of Neural Networks in Financial Forecasting


Financial forecasting plays a crucial role in making informed decisions regarding investments, risk assessment, and general financial planning. With the advancement in technology, the use of artificial intelligence (AI) techniques, particularly neural networks (NN), has gained significant attention in financial forecasting. The objective of this literature review is to analyze and summarize the existing research on the application of NNs in financial forecasting.

Neural Networks and Financial Forecasting

Neural networks are a subfield of AI that have gained immense popularity due to their ability to learn patterns and make predictions based on large datasets. In financial forecasting, NNs are used to model complex relationships between input variables, such as stock prices, interest rates, and economic indicators, to predict future financial outcomes.

One of the early applications of NNs in financial forecasting was in predicting stock market prices. Traditional time series techniques such as autoregressive integrated moving average (ARIMA) models have been used for decades but often struggle to capture the non-linear relationships present in financial data. NNs, on the other hand, excel at capturing these non-linear relationships, making them attractive for predicting stock market prices.

Several studies have demonstrated the effectiveness of NNs in stock price prediction. For instance, Zhang, Patuwo, and Hu (1998) applied a feedforward neural network to predict daily stock market returns. They found that the NN model outperformed traditional time series models in terms of predictive accuracy. Similarly, Kimoto, Asakawa, Yoda, and Takeoka (1990) used NNs to predict daily stock prices and achieved superior results compared to other forecasting methods.

Apart from stock market forecasting, NNs have also been widely used in other financial forecasting tasks, such as foreign exchange rate prediction, bond yield forecasting, and bankruptcy prediction. Huang, Chen, and Chen (2005) conducted a study on forecasting foreign exchange rates using NNs and found that the NN model was able to produce more accurate predictions compared to traditional econometric models. Similarly, Clarke, Harmon, and Nguyen (2011) used NNs to forecast corporate bond yields and reported promising results.

Advancements and Challenges

Over the years, researchers have made several advancements in the application of NNs in financial forecasting. For instance, the use of ensemble methods, where multiple NN models are combined to make predictions, has shown promise. Zhang, Zhou, and Wong (2001) proposed an ensemble approach called an adaptive mixture of local experts (AMLE) to predict stock market indices. They found that the AMLE outperformed individual NN models in terms of prediction accuracy.

In recent years, deep neural networks (DNNs) have garnered significant attention in financial forecasting. DNNs are a subclass of NNs that are capable of automatically learning hierarchical representations of data. This ability allows DNNs to capture intricate features in financial data that may go unnoticed by traditional forecasting techniques. Liu, Ma, and Han (2017) used a deep learning approach to predict the direction of stock price movements and achieved superior results compared to traditional NN models.

Despite the advancements, there are several challenges associated with the application of NNs in financial forecasting. One major challenge is the selection and optimization of NN architectures and model parameters. NNs offer a vast number of design choices, such as the number of hidden layers, the number of neurons per layer, and the activation functions. Selecting the optimal architecture and parameter values requires careful consideration and experimentation.

Another challenge is the requirement for large amounts of data. NNs typically require a significant amount of training data to learn the underlying patterns in financial data. However, obtaining high-quality and diverse financial datasets can be challenging, particularly for emerging or thinly traded markets.


In conclusion, the utilization of neural networks in financial forecasting has shown great promise in capturing complex relationships and making accurate predictions. NNs have been successfully applied in various financial forecasting tasks, including stock market prediction, foreign exchange rate forecasting, and bond yield prediction. Advancements such as ensemble methods and deep neural networks have further enhanced the predictive capabilities of NNs. However, challenges related to model selection and data availability need to be carefully addressed to ensure the reliability and accuracy of NN-based financial forecasting models. Future research in this field should focus on refining NN architectures, exploring alternative training techniques, and expanding the availability of high-quality financial datasets.